Fiber Bragg Gratings for Temperature Monitoring in ...

Fiber Bragg Gratings for Temperature Monitoring in Methanol and Methane Steam

Reformers

By

Élizabeth Trudel

B. A. Sc., University of Ottawa, 2015

A Thesis Submitted in Partial Fulfillment

of the Requirements for the Degree of

MASTER OF APPLIED SCIENCE

in the Department of Mechanical Engineering

© Élizabeth Trudel, 2017

University of Victoria

All rights reserved. This dissertation may not be reproduced in whole or in part, by photocopy or

other means, without the permission of the author.

ii

Fiber Bragg Gratings for Temperature Monitoring in Methanol and Methane Steam reformers

By

Élizabeth Trudel

B. A. Sc., University of Ottawa, 2015

Supervisory Committee

Dr. Peter Wild, Supervisor

Department of Mechanical Engineering

Dr. Ned Djilali, Deparmental Member

Department of Mechanical Engineering

iii

Abstract

Steam reforming of methanol and hydrocarbon are currently the processes of choice to produce

hydrogen. Due to the endothermic nature of these reactions, zones of low temperature are

commonly found in reformers. These zones can potentially damage the reformer through thermal

stresses. Moreover, the response time and size of a reformer are controlled by the heat available to

the reaction. The objective of this thesis is to demonstrate the feasibility of using fiber Bragg

gratings as an alternative solution for temperature monitoring in methanol and methane steam

reformers. To meet this objective, a sensor array containing seven gratings is placed in a metal-

plate test reformer. First, temperature monitoring during methanol steam reforming is conducted

in 12 different sets of conditions. The resulting profile of the temperature change along the length

of the catalyst captures the zones of low temperature caused by the endothermic nature of the

reaction. Several small changes in the temperature profile caused by increasing temperature and/or

flow rates were captured, demonstrating the ability to use these gratings in methanol steam

reforming. Similar experimental work was conducted to validate the possibility of using fiber

Bragg gratings as temperature sensors in methane reforming. Using a regenerated grating array,

data was collected for 13 operating conditions. The conclusions arising from this work are similar

to those drawn from the methanol steam reforming work. The regenerated FBGs exhibited a

behaviour that has not been reported in the literature which is referred to in this thesis as secondary

erasure. This behaviour caused some instability in the grating signal and erroneous readings for

some operating conditions. Despite this, the grating measurements captured the zones of low

temperatures in the reformer and the small changes brought about by increasing the reforming

temperature and lowering the steam to carbon ratio.

iv

Table of Contents

Supervisory Committee .................................................................................................................. ii

Abstract .......................................................................................................................................... iii

List of Tables ................................................................................................................................. vi

List of Figures ............................................................................................................................... vii

Acknowledgments.......................................................................................................................... ix

Chapter 1 - Introduction .................................................................................................................. 1

1.1. Steam Reforming and Current Method for Temperature Monitoring ................................. 3

1.2. Objective ............................................................................................................................ 10

1.3. Thesis Structure ................................................................................................................. 10

Chapter 2 - Literature Review ....................................................................................................... 12

2.1. Fiber Bragg Gratings.......................................................................................................... 12

2.1.1. Operating Principle ..................................................................................................... 16

2.1.2. Applications ................................................................................................................ 18

2.2. Regeneration ...................................................................................................................... 18

2.3. High temperature applications ........................................................................................... 21

2.3.1. Regenerated FBGs ...................................................................................................... 21

2.3.2. Type II FBGs .............................................................................................................. 23

2.4. Summary ............................................................................................................................ 24

Chapter 3 - Experimental Setup .................................................................................................... 25

3.1. Purpose ............................................................................................................................... 25

3.2. Apparatus ........................................................................................................................... 25

3.2.1. Steam Reformer .......................................................................................................... 25

3.2.2. Steam Reforming Test Station .................................................................................... 31

3.3. Experimental Procedure ..................................................................................................... 36

3.3.1. Methanol Steam Reforming Trials.............................................................................. 36

3.3.2. Methane Steam Reforming Trials ............................................................................... 39

3.4. Uncertainty ......................................................................................................................... 41

Chapter 4 - Temperature Monitoring in Methanol Steam Reformers........................................... 43

4.1. Overview ............................................................................................................................ 43

4.2. Characterization ................................................................................................................. 43

4.3. Temperature change in the reformer .................................................................................. 45

v

4.3.1. Temperature profile for each furnace temperature ..................................................... 45

4.3.2. Comparison between the FBG and the thermocouple reference measurements......... 50

4.3.3. Effect of furnace temperature and conversion percentage on the temperature profile 54

4.4. Discussion .......................................................................................................................... 58

Chapter 5 - Temperature Monitoring in Methane Steam Reformers ............................................ 64

5.1. Overview ............................................................................................................................ 64

5.2. Regeneration ...................................................................................................................... 64

5.3. Characterization ................................................................................................................. 67

5.4. Temperature change in the reformer .................................................................................. 70

5.4.1. Temperature drift ........................................................................................................ 70

5.4.2. Temperature profile for each furnace temperature ..................................................... 72

5.4.3. Comparison between the FBG and the thermocouple reference measurements......... 77

5.4.4. Effect of furnace temperature and S/C ratio on the temperature profile ..................... 83

5.4.5. Secondary erasure ....................................................................................................... 86

5.5. Discussion .......................................................................................................................... 88

5.5.1. Outstanding Issues ...................................................................................................... 94

Chapter 6 - Conclusions and Future Work ................................................................................... 97

6.1. Conclusions ........................................................................................................................ 97

6.2. Recommendation for future work ...................................................................................... 99

References ................................................................................................................................... 101

Appendix A – Technical Drawings............................................................................................. 111

Appendix B – Characterization values ....................................................................................... 113

Appendix C – Data analysis ........................................................................................................ 118

Appendix D – Uncertainty .......................................................................................................... 125

Appendix E – Methanol Conversion Percentage ........................................................................ 129

Appendix F – Regeneration ........................................................................................................ 132

Appendix G – Drift Correction Factor ........................................................................................ 137

vi

List of Tables

Table 1 - Summary of the temperature sensing options for steam reformers ................................. 9

Table 2 - Merits and drawback of Type I and Type II FBGs ....................................................... 15

Table 3 – SM125 specifications .................................................................................................... 36

Table 4 – Summary of test conditions for methanol steam reforming trials ............................... 38

Table 5 – Summary of test conditions for methane steam reforming trials .................................. 41

Table 6 – Average characterization values obtained from the five rounds of characterization for

FBG 1-7 ........................................................................................................................................ 44

Table 7 - Methanol conversion percentage calculated for two furnace temperatures, flow rate

=0.110 ml/min ............................................................................................................................... 56


=0.130 ml/min ............................................................................................................................... 57

Table 9 – Average pre-regeneration characterization values of the FBG array ........................... 68

Table 10 – Average temperature sensitivity of the single FBG for different temperature ranges 69

Table 11 - Maximum and minimum temperature change recorded by both FBGs and

thermocouples regardless of flow rates ......................................................................................... 90

vii

List of Figures

Figure 1 - Temperature contours for a tubular methanol steam reformer, wall temperature = 453

K. Reprinted from [26] with permission from Elsevier. ................................................................. 6

Figure 2 - Schematic of a top-fire methanol steam reformer with multiple reforming chambers.

Adapted from [27] with permission from Wiley. ........................................................................... 7

Figure 3 - Schematic of an FBG. .................................................................................................. 13

Figure 4 – Top view of the base section of the reformer including the catalyst plate and the

Thermiculite strip. ......................................................................................................................... 26

Figure 5- Section view of the inlet of the reformer showing the holes to accommodate the fiber.

....................................................................................................................................................... 27

Figure 6 – Photograph of gas inlet side of the furnace with thermocouple and fiber inlet into the

furnace........................................................................................................................................... 28

Figure 7 – Top view of the reformer showing the position of the thermocouples and FBGs. ..... 29

Figure 8 – Section view of the reformer showing the position of a thermocouples and the fiber

with respect to the catalyst plate and the reaction site. ................................................................. 30

Figure 9- Thermocouples imbedded in the reformer base. ........................................................... 30

Figure 10 – Schematic of the steam reforming test station. .......................................................... 32

Figure 11 – Reformer installed in the furnace before steam reforming trials. .............................. 33

Figure 12 – Screen capture of the LabVIEW interface. ................................................................ 35

Figure 13 - Temperature change in the reformer as a function of position for a furnace

temperature of 250 °C. .................................................................................................................. 46


temperature of 265 °C. .................................................................................................................. 47


temperature of 280 °C. .................................................................................................................. 48

Figure 16 - Temperature change as a function of position in the reformer for a furnace

temperature of 250 °C measured by FBGs and thermocouples. ................................................... 51



Figure 18- Temperature change as a function of position in the reformer for a furnace


Figure 19 - Temperature change as a function of position in the reformer for a flow rate of 0.110

ml/min. .......................................................................................................................................... 55


ml/min. .......................................................................................................................................... 57

Figure 21 - Photograph of the inside of the reformer after the methanol steam reforming trials. 59

Figure 22 - Power and temperature over time during the FBG regeneration. .............................. 65

Figure 23 - Full spectrum of the FBG array before and after regeneration. ................................. 66

Figure 24- Wavelength over time of FBG 1 during experiments at 650 °C showing the

temperature drift inside the reformer ............................................................................................ 71


temperature of 650 °C. .................................................................................................................. 73

viii


temperature of 750 °C. .................................................................................................................. 74


temperature of 775 °C. .................................................................................................................. 76







Figure 31 - Temperature change in the reformer as a function of positions for two different

furnace temperatures and S/C ratio (a)- flow rate of 40 ml/min (b)- flow rate of 60 ml/min....... 84

Figure 32 - Full spectrum of the FBG array undergoing secondary erasure while the furnace

temperature is increasing from 775 °C to 825 °C. ........................................................................ 87

Figure 33 - Full spectrum of the FBG array showing the change in the noise floor level. ........... 95

ix

Acknowledgments

I would like to acknowledge the many people who have contributed in one way or another

to this work. First, I would like to thank my supervisor Dr. Peter Wild for guiding me through this

work and providing me with the tools to grow as an engineer and a person. I would also like to

express my gratitude to Dr. Brant Peppley for his guidance on the steam reforming aspect of this

work and for the opportunity to tackle this project. I would like to thank Graeme Clancy and Aidu

Qi from the Fuel Cell Research Centre for their help preparing and running the experiments. Thank

you to Elise and Erica for always welcoming me in Kingston with open arms. I would like to thank

my colleagues and friends from the University of Victoria; Reza Harirforoush, Geoff Burton and

Dr. Luis Melo, it has been a pleasure working alongside you. Thank you to my parents and sister

for always encouraging me to pursue my goals. And finally, thank you Jeff for believing in me.

1

Chapter 1 - Introduction

Global growth in energy consumption over the past few decades, primarily met by using fossil

fuels, has caused large amounts of greenhouse gases (GHGs) to be released into the atmosphere

[1]. These GHGs are major drivers of climate change. Renewable energy sources, such as wind,

solar and hydro are increasingly used as alternatives to fossil fuels for electricity generation. A key

issue arising from the use of such renewable energy sources is their intermittency. Electricity can

only be generated when the resource is available. Given the cost of current energy storage

solutions, this is not yet a feasible solution to resource intermittency [2]. A hydrogen economy, in

which energy is delivered through the hydrogen currency, has been proposed to mitigate the effect

of climate change by providing a solution to the intermittency of renewable energy sources [3]–

[5]. In this scenario, hydrogen is produced through electrolysing of water or steam reforming when

energy production exceeds energy demand. The extra energy produced is used to power the

electrolyser or used to generate the heat required by the steam reforming process.

The transportation sector, which is heavily reliant on oil, is ranked globally as the fourth

largest emitters of GHG among the key economic sectors. In Canada, the transportation sector is

the second largest emitter of GHGs after the oil and gas industry [6], [7]. Using clean hydrogen in

fuel cells is one of the options currently being proposed to reduce the GHG emissions caused by

the transportation sector [8]–[10]. Clean hydrogen is defined as hydrogen being produced from a

process which releases no GHG in the atmosphere. Clean hydrogen can, for example, be obtained

2

from electrolysis of water powered by renewable energy sources or from steam reforming of

hydrocarbons in conjunction with carbon storage and sequestration (CSS).

Hydrogen is the third most abundant element in the atmosphere. Unfortunately, hydrogen is

not found naturally as a gas. It must be produced from another source. Therefore, hydrogen is not

a primary fuel but rather an energy carrier or currency. There exist multiple ways to produce

hydrogen, such as steam reforming of hydrocarbons or methanol, electrolysis of water, partial

oxidation, gasification and solar thermochemical splitting of water just to name a few. The cost of

hydrogen production will vary by region based on electricity and feedstock price. But according

to recent studies, the cost of hydrogen produced through electrolysis is currently much higher than

the cost of hydrogen produced through steam reforming which gives the latter an edge as the

process of choice for hydrogen production [8], [11], [12]. Moreover, the price of other hydrogen

production method such as solar thermochemical splitting of water is expected to decrease as

further research and development progress.

Recent studies suggest that, if the transportation sector shifts from oil to hydrogen as its

primary fuel, the majority of the hydrogen produced in 2030 will still come from fossil fuels,

mainly natural gas [9], [13], [14]. This potential intensification of hydrogen production to meet

growing demand associated with the increased use of fuel cells, while maintaining the current

industrial demand, highlights the importance of improving the efficiency of the steam reforming

process.

3

1.1. Steam Reforming and Current Method for Temperature Monitoring

Steam reforming is a chemical reaction in which a fuel reacts with steam over a catalyst to

produce hydrogen and carbon dioxide. Steam reforming takes place at high temperature and can

produce carbon monoxide as a by-product when incomplete conversion occurs. A variety of

hydrocarbons can be used as fuel for steam reforming but, for many reasons, methane (CH4) is

preferred.

Methane is the primary component in natural gas, ranging from 87% to 97% on a molar

basis [15]. Canada is the fifth largest producer of natural gas in the world and natural gas is found

in abundance in many provinces in Canada, such as British Columbia, Alberta and Saskatchewan

[16]. Moreover, methane steam reforming takes advantage of the natural gas infrastructure in place

across Canada, the United States and Europe which could facilitate the hydrogen production and

methane delivery processes.

Methane has a low carbon to hydrogen ratio and a lack of strong carbon-carbon bonds.

Carbon-carbon bonds are very strong, and therefore, more energy is required to break this type of

chemical bond. Natural gas steam reforming, or methane steam reforming, is currently the source

of 95% of the hydrogen production in the United States and 48% of the global hydrogen production

[13], [17]. The current price of natural gas, 3.30 USD/MMBtu, gives it an advantage compared to

other feedstocks [18]. The chemical reaction for methane steam reforming is presented below in

Equation (1.1) and takes place between 600 °C and 850 °C.

4

CH4 + 2H2O → CO2 + 4H2 ∆H2980 = 165 kJ/mol (1.1)

Methanol steam reforming is also a high temperature chemical reaction between the fuel,

methanol, and steam over a catalyst. This reaction takes place at much lower temperatures,

between 230 °C and 300°C, than methane steam reforming. Methanol is an alcohol which is liquid

at ambient temperature and thus simplifies the transportation and storage issues associated with

gases such as methane or hydrogen [19]. The methanol steam reforming chemical reaction is as

follows:

CH3OH + H2O → CO2 + 3H2 ∆H2980 = 49.4 kJ/mol (1.2)

The effect of temperature on the steam reforming process has been studied extensively [20]–

[23]. In methane steam reforming, studies have found that the hydrogen yield increases with

increasing temperature [20], [23]. Increasing the reforming temperature also increases the carbon

monoxide in the products [21]. The water gas shift (WGS) reaction, Equation (1.3), is

thermodynamically advantaged at lower temperatures while the reverse water gas shift (RWGS)

reaction, which produces carbon monoxide, is advantaged at high temperatures.

Not only is carbon monoxide a highly-regulated pollutant but small concentrations of this

molecule are poisonous to a proton exchange membrane fuel cell (PEMFC). The generation of

5

carbon monoxide can also lead to the creation of solid carbon through the Boudouard reaction,

shown in Equation (1.4). The deposition of solid carbon on the catalyst layer leads to its

deactivation [20]. Sufficient heat must be provided to the reaction to ensure a worthwhile hydrogen

yield and fuel conversion percentage but the temperature must be controlled within a narrow range

to minimize formation of carbon monoxide which can lead to premature catalyst deactivation.

Water gas shit reaction: CO + H2O ↔ CO2 + H2 (1.3)

Boudouard reaction: 2CO ↔ CO2 + C (1.4)

Steam reformers must, therefore, have good heat transfer characteristics. Monitoring the

temperature at or around the reaction site can be used to identify areas of improvement for heat

transfer. If a significant region of low temperature is found near the reaction site, it indicates an

area of poor heat transfer. As seen in Equations (1.1) and (1.2), both methane and methanol steam

reforming are endothermic reactions. Therefore, a temperature drop at the inlet of a reactor is

unavoidable. This temperature drop occurs when the reactants first encounter the catalyst and start

reacting. The reaction absorbs heat from its surroundings, creating a zone of low temperature.

After the steam reforming reaction has taken place, the product flows through the rest of the

reaction chamber and the temperature will increase and return to the furnace temperature, also

defined as reforming temperature. The modeling of a tubular reformer done by Ribeirinha et al.

shows a temperature drop of almost 7 K at the reformer inlet compared to the wall temperature

[24]. The results of this modeling are pictured in Figure 1.

6

Figure 1 - Temperature contours for a tubular methanol steam reformer, wall temperature = 453

K. Reprinted from [24] with permission from Elsevier.

While this zone of low temperature is to be expected due to the endothermic nature of the

chemical reaction, its magnitude should be minimized. In a commercial reformer with multiple

reforming chambers, temperature discrepancies between reaction chambers at a given distance

from the inlet can occur due to poor heat transfer within the reformer. These temperature

discrepancies can cause significant damage to the reformer, caused by thermal stresses, and lower

the thermal efficiency of the plant [25]. An example of a top-fired methanol steam reformer with

multiple reforming chambers is illustrated in Figure 2. For a reformer coupled with a fuel cell

system, the response time (i.e. the time for a reformer to react to a change in the demand for

hydrogen) is controlled by the heat available to the reaction. The response time directly affects

the ability of the reformer to follow the dynamic load demand of the fuel cell [26]–[28].

7

Figure 2 - Schematic of a top-fire methanol steam reformer with multiple reforming chambers.

Adapted from [25] with permission from Wiley.

Currently, temperature monitoring of reformers is done using thermocouples, resistance

temperature detectors (RTDs) or thermal imagers [25], [27], [29]–[31]. While these measurement

methods are all well-established, they are not suitable for use in steam reformers.

While thermocouples have been used successfully for decades, each thermocouple is a

stand-alone probe with a two-wire connection. The most commonly used thermocouple is the Type

K. Thermocouple size varies based on the application but they can be as small as 3 mm in diameter.

To obtain high-resolution temperature measurement throughout the reformer requires a large

number of thermocouples and the associated wiring can be cumbersome. This issue limits the

number of sensors that can be used, based on the design of the reformer.

8

RTDs are relatively small and have been used successfully in methanol steam reforming

[32]. A thin-film RTD can be as small as 4 mm2 while wire-round RTDs are comparable in size to

thermocouples. On the other hand, the maximum operating temperature of a commercially

available RTD is less than 600 °C which is below the minimum reforming temperature for methane

steam reforming. This can be increased to 750-850 °C for a custom order which is still lower than

the temperature of some methane steam reformers [33]. Moreover, these sensors each require two

or three lead-in wires which can also limit the number of sensors used depending upon the design

of the reformer.

Finally, thermal imagers can provide a simple and cost-effective alternative for temperature

monitoring in steam reformers. Thermal imagers can help to quickly identify zones of low

temperature but only at the surface of the reformer which is a drawback for this application.

Unfortunately, thermal imagers are not as accurate as other types of temperature sensors. The

reported accuracy of a thermal imagers is ±2-3% or ±3°C which ever one is largest [34]. Moreover,

in situation where the reformer must be located in an enclosed area for heating purposes, it is not

possible to obtain temperature data directly at, or around, the reaction site which is where zones of

low temperature would occur.

An alternative approach for temperature monitoring in steam reformers is fiber optics sensors

(FOS) based on the fiber Bragg grating (FBG). While there exist other types of FOS such as long

period gratings and Fabry Perot interferometers, these are not commonly used for temperature

sensing and will not be discussed here. Rayleigh backscattering fiber sensors are sometimes used

9

as temperature sensors [35] but the cost of this system is prohibitive and thus, this type of FOS

will not be discussed here either. FBGs are off-the-shelf, well established devices that can be used

to measure temperature and/or strain [36], [37]. Immunity to electromagnetic interference,

multiplexing, small size and high sensitivity are some of the characteristics that make FBGs

advantageous compared to other types of sensors. Table 1 summarises the characteristics, spatial

resolution and temperature resolution, as well as the merits and drawbacks of the four types of

sensors described in this section.

Table 1 - Summary of the temperature sensing options for steam reformers

Spatial

resolution

Temperatur

e resolution

Accurac

y

Merits Drawbacks

Type K

thermocouple

s

Will vary

based on the

probe size

and design

configuratio

n

0.1 °C [38] ±2.2 °C

or ±0.75

% which

ever one

is greater

[39]

Mass

manufactured,

low cost, high

operating

temperature

Non-

multiplexing,

prone to error

caused by

electromagneti

c interference,

requires one

access point to

the reformer

per sensor

RTDs Will vary

based on

small and

configuratio

n

0.01 °C [40] 0.03 °C

to 0.3 °C

[41]

Small size,

multiple

configurations

possible (thin-

film, wire-

round)

Non-

multiplexing,

fragile,

expensive,

slow thermal

response [41],

prone to error

caused by

electromagneti

c interference,

requires one

access point to

the reformer

per sensor

Thermal

imagers

N/A 1 °C [33],

[34]

±3 °C or ±2-3%

which

ever one

Low cost, easy

to use

Low maximum

working

temperature, no

access to the

10

is greater

[34]

inside of the

reformer [33]

FBGs 1 mm [42] 0.1 °C [43] 0.1 °C Multiplexing,

immunity to

electromagneti

c interference,

requires only

one access

point in the

reformer

Brittle, cross-

sensitivity to

strain,

expensive

1.2. Objective

The objective of the work presented in this thesis is to demonstrate the feasibility of using

FBGs for temperature monitoring in a steam reformer.

A single channel metal plate reformer is modified to enable installation of FBG temperature

sensors adjacent to the flow path. The sensors are used to monitor temperature profiles at various

reforming temperatures and flow rates in both methanol and methane steam reforming. To the best

of the author’s knowledge, FBGs have not been previously used for temperature monitoring in

steam reformers and, therefore, this work presents a novel application of FBGs.

1.3. Thesis Structure

The thesis is structured as follows: Chapter 2 includes a thorough review of the literature

on FBGs, their operating principle, advantages and limitations. This chapter also covers instances

where FBGs have been used as temperature sensors in high temperature environments. Chapter 3

11

lays out details concerning the experimental procedure which includes the experimental set-up, the

methods for the methanol and methane steam reforming trials.

Chapters 4 and 5 highlight the experimental results for methanol and methane steam

reforming, respectively. This includes sensor characterization, regeneration in the case of methane

steam reforming, temperature monitoring trial results, data analysis and relevant discussion

associated with the results and their relevance to the field of hydrogen production through steam

reforming.

Chapter 6 concludes this thesis with recommendations for future work as well as a brief

conclusion concerning the contributions arising from this work.

12

Chapter 2 - Literature Review

2.1. Fiber Bragg Gratings

Over the past few decades, fiber optic sensors (FOS) have been developed for many

applications. This trend is associated with the increasing use of fiber optics in the

telecommunication sector. Among the most common FOS are those based on fiber Bragg gratings

(FBGs). An FBG is a region of periodic modulation of the refractive index of the core of an optical

fiber. An FBG can vary in length from 1 mm to 24 mm, and has a submicron period [35]. An FBG

enables coupling of the fundamental core mode and the counter-propagating core mode [44]. When

an FBG is exposed to a broadband light source, it acts as a stopband filter reflecting a narrow

wavelength range. The peak of this reflected range is at the Bragg wavelength, λB. Figure 3 presents

a schematic of the FBG operating principle.

As mentioned briefly in Chapter 1, FBGs are off-the-shelf devices and can be used to

measure temperature and/or strain [36], [37]. Immunity to electromagnetic interference,

multiplexing, small size and high sensitivity are some of the characteristics that make FBGs

advantageous compared to other types of sensors.

13

Figure 3 - Schematic of an FBG.

FBGs are classified by their manufacturing, or writing, process and referred to as Type I,

Type II or regenerated FBGs. This classification is important since the writing process will greatly

affect some of the characteristics of the FBG. The manufacturing process and characteristics of

Type I and Type II FBGs are discussed below while regenerated gratings are the subject of Section

2.2.

Type I FBGs are the oldest and most commonly used FBGs. A Type I FBG results from a

periodic modulation of the refractive index (RI) of the core of a germanium-doped optical fiber by

exposure to UV light, generally 244 nm, through a phase-mask [36]. In some instances, the fiber

14

is loaded with hydrogen at high pressure to enhance photosensitivity [45]. Type I FBGs are

characterized by high reflectivity which can approach ~100 % [36]. This type of FBG is not

suitable for applications where the temperature can reach upwards of 450 °C [46]. In such

environments, the periodic modulation of the RI will erase. Regeneration of type I FBG increases

their operating temperatures to ~1295 °C [47]. Regeneration will be discussed in Section 2.2.

FBGs classified as Type II are also commonly referred to as damage gratings. They are

written with a high peak power ultraviolet laser which results in a permanent damage to the fiber.

It is possible to manufacture Type II FBGs from a single laser pulse and, in some instances, this

can be done directly as the fiber is being drawn from the tower [48]. While Type II FBGs have

certain advantages over Type I FBGs, such as resistance to temperatures up to 1000 °C, the nature

of their manufacturing process results in high scattering losses and reduced mechanical strength

of the fiber [46], [49]. High scattering loses can reduce the number of FBG that can be multiplexed

on one array [50].

While most Type I FBGs are written in germanium-doped silica fiber or in boron and

germanium-codoped silica fiber, Type II FBGs can be written in pure silica or sapphire fiber. The

advantages of pure silica fiber over boron and/or germanium-doped fiber is related to hydrogen

darkening of the fiber. Hydrogen darkening can be either permanent or reversible [51]. Reversible

hydrogen darkening is caused by molecular hydrogen infiltrating the fibre structure and affecting

its optical properties. Hydrogen loss increases with an increase in hydrogen partial pressure. This

is referred to as reversible hydrogen as the losses will be reverted once the fiber is removed from

15

the hydrogen environment. Reversible hydrogen darkening can affect any type of silica-based

optical fiber. Permanent hydrogen darkening is caused by the reaction between the hydrogen

molecule and germanium defect sites within the core of the optical fiber [51]. The magnitude of

these losses is proportional to partial pressure of hydrogen but, unlike reversible darkening, it is

dependent on temperature.

Recently, fiber optic sensors, such as Type II FBGs, have been written in sapphire fiber.

Sapphire fiber can be used at temperatures of up to 1745 °C and has been proven to be stable for

28 days at 1400 °C. This type of fiber is immune to hydrogen darkening. Some of the disadvantages

of using sapphire fiber include its cost and the multimode nature of the fiber which leads to high

optical attenuation. In addition, splicing sapphire fiber to regular silica fiber results in high losses

[52].

In steam reforming, it is desirable to measure the temperature profile with a high resolution

to identify areas of low temperature. For this reason, Type II FBGs are not suitable for the

application and type I FBG are selected as the sensor of choice for this application. Table 2

summarises the characteristics of Type I and Type II FBGs as discussed here.

Table 2 - Merits and drawback of Type I and Type II FBGs

Merits Drawbacks

Type I FBGs Off the shelf devices

Resist temperatures of up to

1295 °C after regeneration

Susceptible to both

reversible and permanent

hydrogen darkening

16

Regeneration needed to

sustain temperatures above

450 °C

Type II FBGs No regeneration needed to

sustain high temperatures

Can be written in sapphire

fiber which is immune to

hydrogen darkening

High scattering loses

limiting the number of

sensors per array

Manufacturing process

reduces mechanical

strength of the fiber

Susceptible to both

reversible and permanent

hydrogen darkening if

written in standard

telecommunication grade

fiber

2.1.1. Operating Principle

As mentioned earlier, an FBG can be used to measure temperature and/or strain. When

using FBGs to measure these properties, the information concerning the measurand is encoded in

the Bragg wavelength. The Bragg wavelength of an FBG is related to the effective refractive index

of the fiber, neff, and the period of the grating, Λ, as shown in Equation (2.1).

𝜆𝐵 = 2𝑛𝑒𝑓𝑓Λ (2.1)

When a change in temperature occurs, there is a change in the refractive index as well as a

change in the period of the grating caused by thermal expansion of the fiber. The changes in these

two properties are captured in the thermo-optic coefficient, ξ, in Equation (2.2). Authors in the

literature have reported temperature sensitivity of FBGs ranging 10 to 13 pm/°C for a Bragg

wavelength of ~1550 nm [36], [44]. Although other authors have reported that this temperature

17

sensitivity is valid only for a temperature range of approximately 0 to 80 °C [44]. As the

temperature increases, so does the temperature sensitivity of the FBG. Some authors have reported

a sensitivity of ~15 pm/°C for temperatures ranging from 200 to 500 °C [49].

Δ𝜆𝐵 = 𝜆𝐵(1 + 𝜉)Δ𝑇 (2.2)

FBGs are also commonly used as strain sensors in a variety of applications. Like an FBG

temperature sensor, the information concerning this measurand is encoded within the wavelength.

The strain sensitivity of FBGs has been reported in the literature around ~1.2 pm/µε [36]. For this

measurand, the shift in Bragg wavelength is caused by the change in the period due to the physical

elongation induced by strain as well as the changes in the refractive index due to the photoelastic

effects [53]. This is expressed by the photoelastic coefficient, ρa. The change in Bragg wavelength

induced by a change in strain is presented below in Equation (2.3). The combined effect of strain

and temperature on the shift in Bragg wavelength is presented in Equation (2.4).

Δ𝜆𝐵 = 𝜆𝐵(1 − 𝜌𝑎)Δ𝜀 (2.3)

Δ𝜆𝐵 = 𝜆𝐵(1 + 𝜉)Δ𝑇 + 𝜆𝐵(1 − 𝜌𝑎)Δ𝜀 (2.4)

18

2.1.2. Applications

FBGs are used as sensors in numerous applications. One of the most common applications

for FBGs is strain sensor in structural health monitoring where the fiber is inserted within the

structure to monitor the stresses applied to it [54]–[56]. In some instances, one of the FBG on the

structure is strain isolated and used to measure temperature, this temperature value is then used to

compensate for temperature effects on the Bragg wavelength [57]. There are also reports in the

literature of FBGs used as pressure and temperature sensors in downhole environment in the oil

and gas industry [58], [59]. Another instance of FBGs used as strain sensors is the highly sensitive

catheter presented by Bueley and Wild [60]. In this article, one of FBGs is also used to measure

temperature as this information is needed to compensate for the temperature sensitivity of the

sensors. David et al. reported on the use of FBGs as temperature sensors in a polymer electrolyte

membrane fuel cell (PEMFC) [61]. David et al. have also reported on the simultaneous

measurement of relativity humidity (RH) and temperature in a PEMFC using optical sensors based

on FBGs [62]. A more in-depth look at the use of FBGs in high-temperature environment is

presented in Section 2.3.

2.2. Regeneration

As mentioned earlier, Type I FBGs can only be used in environments where the

temperature is below 450 °C [46]. Above this temperature, the degradation of the FBG, which is

caused by thermal decay, is too great and eventually results in its complete erasure. The various

advantages of Type I FBGs over other types of sensors, including Type II FBGs, has driven

19

researchers to look for ways to increase the operating temperature of Type I FBGs while retaining

their advantageous characteristics. Once such way is the regeneration of Type I FBGs.

Regenerated FBGs are obtained from Type I FBGs which undergo high-temperature treatment.

The Type I FBGs used to produce regenerated grating are called seed FBGs. To obtain regenerated

FBGs, the seed FBGs must be heated up to a temperature of approximately ~900-950 °C [63],

[64]. At this temperature, a rapid decrease in seed FBG’s reflectivity, which is the first step of the

regeneration process, can be observed. This is followed by complete erasure of the seed FBG

which is then accompanied by the growth of the regenerated FBG. The growth of the regenerated

FBG is manifested by its increase in reflectivity at or near the seed FBG’s Bragg wavelength [63].

Through the development of regeneration, regenerated FBGs can now be used at temperatures of

up to 1295°C, which approaches the glass transition temperature [47], [64].

Regenerated gratings were first reported in 2008 by Canning et al. and Bandyopadhyay et

al. [47], [64]. Since then, more work has been done to try to understand the theory behind

regeneration. While the exact theory behind regenerated FBGs is not yet known, some authors

have published theories based on their observations. Among these, Bandyopadhyay et al. suggests

that the manufacturing process of Type I FBG leaves a signature in the fiber, most likely at the

core-cladding interface or within the inner cladding itself, which is not erased through the

regeneration procedures detailed previously [64]. Other authors have reported that regenerated

FBGs do not occur in a fiber which has not undergone hydrogen loading as part of the writing

process [64], [65]. Lindner et al. discovered that hydrogen loading during the writing was not

20

necessary for regeneration and that obtaining regenerated FBGs was still possible if the fiber was

loaded with hydrogen during the thermal processing [66].These finding highlights the importance

of hydrogen during the regeneration process but what exact mechanism is essential to obtain

regenerated FBGs is still unclear [46].

Bandyopadhyay et al. have studied the effect of different regeneration schedules on the

strength of the regenerated FBG [67]. In this article, they used three single FBGs fabricated in an

identical manner and subjected to different regeneration schedules. They found that the stronger

regenerated FBG was annealed isothermally at the regeneration temperature. The FBG was heated

to 900 °C, the erasing temperature, in 60 minutes and kept constant for 20 minutes. After

regeneration, the temperature was raised to 1050 °C to allow the regenerated FBG to stabilize.

Another example of the regeneration procedure commonly used in the literature is as follows:

initial temperature set to 600 °C and reached in 10 minutes followed by steps of 50 °C set every

10 minutes until 1100 °C is reached [47], [63], [64].

Although the theory of regeneration remains elusive, this process has been used

successfully on numerous occasion to create sensors based on regenerated FBGs. The following

section highlights some of the literature on using regenerated FBGs as sensors in high-temperature

environment.

21

2.3. High temperature applications

2.3.1. Regenerated FBGs

The development of regenerated FBGs has opened a new window of opportunity for fiber

optics sensors based on FBGs in high-temperature environments. Despite the relative novelty of

regenerated FBGs compared to other types of temperature sensors, FBGs have been used in a

number of applications due to their advantages.

The use of multiplexed regenerated FBGs was demonstrated by Laffont et al. in their work

on mapping of the temperature gradient in a tubular furnace [68]. In this work, the seed FBGs are

written in Corning SMF-28 fiber using the phase mask method. The FGBs are regenerated

simultaneously, as follow: heat up to 710 °C and pre-anneal at this temperature for 2 hours,

increase temperature to 920 °C and maintain for 1 hour followed by a passive return to room

temperature. The array’s temperature sensitivity was then characterized between 50 and 900 °C

before being used to map the temperature gradient of a vertical tubular furnace. In order to protect

the fiber throughout the experiments, a metallic packaging was applied to the fiber prior to

regeneration. Laffont et al. also published the results of their work on using multiplexed

regenerated FBGs as temperature sensors in a sodium-cooled fast reactor [69]. This application

requires sensors at a sustained temperature above 550 °C. Multiplexed regenerated FBGs were the

sensor of choice for this work and the regeneration schedule followed was the same as described

for [68]. Using multiplexed regenerated FBGs allows precise temperature mapping inside the

reactor core. This work validated the use of regenerated FBGs in liquid sodium heated up to 500

°C. The authors also evaluated the long-term stability of the multiplexed regenerated FBGs by

22

placing them at high temperature for a period of 2000 hours. This work shows an initial decrease

in reflectivity followed by its stabilization which is maintained through the remainder of the

experiment.

The work of Barrera et al., while not describing an application where FBGs where use in a

high temperature environment, provide some good insight on the performance of regenerated

FBGs as high-temperature sensors [70], [71]. The temperature treatment required to obtain

regenerated FGBs is known to render the optical fiber quite brittle. If the sensors will be moved or

manipulated after regeneration, a packaging is necessary to avoid manipulation of the bare fiber.

In their work, Barrera et al. analyze the performance of FBGs with or without packaging. This is

done by measuring the temperature sensitivity, the response time and possible hysteresis in the

thermal response. The packaging used in both articles is described as a two-bore ceramic casing

in which the optical fiber sits. The ceramic casing is inserted in a metal casing, INCONEL 600

nickel alloy, with an external diameter of 1.5 mm. The response time measured by the authors is

of 9 seconds for a temperature increase of 1000 °C and 22 seconds for a decrease in temperature

of the same magnitude. The response time for a temperature increase is said to be similar to the

one of a commercial thermocouple. Temperature cycling tests are performed to measure the

possible hysteresis of the sensor and the authors concluded that there is no observable hysteresis

for a packaged FBG. Furthermore, there is no sensor degradation due to repeated exposure to high

temperature during the temperature cycling test. Finally, both packaged and unpackaged

regenerated FBGs have the same thermal sensitivity. The results of their work prove that the

packaging of an FBG enhances its mechanical strength without affecting its performance as a

23

temperature sensor. It also demonstrates the feasibility of using such sensors to measure

temperature in harsh environments.

2.3.2. Type II FBGs

Due to their high-temperature resistant nature, Type II FBGs have also been used on

multiple occasions for high-temperature sensing. Walker et al. present the development and testing

of four sensor arrays each comprising of 21 Type II FBGs used to monitor the temperature

gradients on the sidewall and exhaust of a gas turbine combustor simulator [72]. The FBG arrays

were protected using a packaging made of standard 316 stainless steel. Using the Type II FBGs,

they created temperature maps of the sidewall and exhaust of the gas turbine combustor simulator.

These results were compared and in good agreement with the Type N thermocouple reference

measurements. This work builds upon previous work by Willsch et al. who evaluated a variety of

FOS, including Type I, Type II and regenerated FBGs, for use as temperature sensors in gas turbine

monitoring [73]. Initial testing was performed with 60 Type II FBGs installed in a honeycomb

fashion in the exhaust path of gas turbine. According to Willsch et al., these FBG arrays operated

for 8 months at 600 °C.

Another example of successfully using Type II FBGs at high temperature was published

by Black et al. [74]. In this work, the authors demonstrated the possibility of using two arrays,

totaling 8 FBGs, for temperature mapping of a thermal protection system (TPS). A TPS is a heat

shield found on spacecraft performing atmosphere entry or re-entry. A temperature monitoring

system could help reduce risks and allow TPS mass reduction. Due to the unique environment in

24

which TPS operates, there is a possibility of tensile load being applied to the sensors. The authors

demonstrated the ability of the Type II FBG arrays to survive up to 1000 °C and the ability of the

array to be loaded in tension while also being subjected to temperatures of up to 1000 °C. The

tensile force applied to the array was 6.8 kpsi and no breakage occurred.

2.4. Summary

This chapter provides an overview of the operating principle and applications of type I and

type II FBGs. The merits and drawbacks of using each type of sensor are summarised in Table 2.

As described in Section 2.3, both types of FBGs can operate in the environmental conditions of

methanol and methane steam reformers and could be suitable temperature sensor options for these

applications. A high-resolution temperature profile is desired to identify zones of low temperature

in a reformer. To achieve this, multiple FBGs need to be multiplexed on one array which could

become an issue with Type II FBGs. Moreover, Type I FBGs are currently commercially available

and have been proven to work at temperatures of up to 1295 °C when regenerated which is much

higher than the methane steam reforming temperature. Therefore, type I FBGs are the sensor of

choice for temperature monitoring in steam reformers. Chapter 3 will introduce and detail the

methodology applied to the use of Type I FBGs as temperature sensors in methanol and methane

steam reformers.

25

Chapter 3 - Experimental Setup

3.1. Purpose

The purpose of the experiments described in this chapter is to prove the feasibility of using

FBGs as temperature sensors in methanol and methane steam reformers. As mentioned earlier, this

represents a novel application for fiber optic sensors based on FBGs. To accomplish this, an

existing steam reformer is modified to accommodate the fiber. This chapter also describes the

steam reforming test station where experiments took place. Since the objective of this thesis is to

demonstrate the feasibility of using FBGs as temperature sensors in methanol and methane steam

reformers, two different trials are described in Section 3.3. The two trials are, firstly, methanol

steam reforming and, secondly, methane steam reforming.

3.2. Apparatus

3.2.1. Steam Reformer

An existing reformer previously developed by Aida Khosravi to study reaction kinetics on

catalytically coated heat transfer components was used throughout this work [75]. The reformer

was modified to accommodate the fiber on which the FBGs are written. Moreover, other

modifications were made to prevent gaseous chemical leakage from the reaction chamber during

steam reforming experiments. A top view of this reformer, including the catalyst plate and sealing

Thermiculite gasket (Flexitallic Group, USA) are shown in Figure 4. Thermiculite 866 was

selected as a gasket material by Aida Khosravi for her work with this steam reformer due to its

26

sealing properties during hydrocarbon steam reforming [75]. According to its manufacturer,

Thermiculite 866 is the sealing material of choice for SOFC. Thermiculite experiences no

reduction in material thickness at high temperature, ensuring a seal that is maintained in service,

and no burn off of organic material at high temperature [76]. Moreover, Thermiculite 866 can

sustain temperatures of up to 1000 °C which is satisfactory for this application.

Figure 4 – Top view of the base section of the reformer including the catalyst plate and the

Thermiculite strip.

To accommodate the fiber, a hole was end-milled on the gas inlet side of the reformer, this

is pictured in Figure 5. The hole was 1/8 inch in diameter and was drilled through a depth of 3/8

in. A smaller hole, diameter of 1/16 inch, was extended from there through to the pocket onto

which the catalyst plate sits. The bottom of this second hole coincide with the bottom of the groove

described in the next paragraph. This was designed in order to support the fiber. The technical

27

drawings showing a section view of the inlet of the reformer and the modifications to accommodate

the fiber described in this paragraph can be found in Appendix A.

Figure 5- Section view of the inlet of the reformer showing the holes to accommodate the fiber.

A groove was milled along the bottom of the pocket where the catalyst plate sits, this is

pictured above in Figure 5. This groove had a depth of 1/16 inch and a width of 1/16 inch. It runs

along the length of the reformer through to the end of the pocket. The purpose of this groove is to

protect the fiber from the strain that would be induced if it were to be placed directly between the

catalyst plate and the surface of the pocket in which sits the catalyst plate. By placing the fiber in

this groove, it is isolated from strain due to interaction with the reformer so that changes in Bragg

wavelength will be due solely to changes in temperature. External to the reformer, the fiber is

protected by a stainless steel tube with a diameter of 1/16 inch which can be seen in Figure 6. This

28

tube elongates past the hot zone where it can be sealed with PEEK tubing and Yor-Lok (Parker

Hannifin, USA) compression fittings.

Figure 6 – Photograph of gas inlet side of the furnace with thermocouple and fiber inlet into the

furnace.

The fiber is placed in this groove and is sealed from the reformer gases, which include

hydrogen, by a Thermiculite gasket. The Thermiculite gasket is cut by hand to match the flow

channel and geometry of the reformer. It is then placed between the top and bottom pieces of the

reformer and prevents the gases from travelling out of the reformer and under the catalyst plate.

The Thermiculite gasket can be seen in Figure 4.

The reformer was designed with holes in the base, as shown in Figure 8, to insert seven

Type K thermocouples. These thermocouples are used throughout the experiments to provide

29

reference measurements. In previous experimental work done using this reformer, some issues

with sealing occurred [75]. To ensure that an effective seal is maintained, copper plugs were

manufactured and press fit into the existing holes. These can be seen in Figure 7 and Figure 8. The

thermocouples are embedded under the reformer and sit 6.35 mm below the of the bottom surface

of the catalyst plate. The holes in which they are inserted is pictured in Figure 8 while Figure 9

illustrates how the thermocouples would be positioned when the reformer is placed in the furnace.

Figure 7 offers a top view of the reformer showing the position of the thermocouples and

FBGs. Figure 8 gives a different perspective and offers a section view of the reformer which helps

visualize the position of the thermocouples with respect to the FBGs and the catalyst plate.

Figure 7 – Top view of the reformer showing the position of the thermocouples and FBGs.

30

Figure 8 – Section view of the reformer showing the position of a thermocouples and the fiber

with respect to the catalyst plate and the reaction site.

Figure 9- Thermocouples imbedded in the reformer base.

31

3.2.2. Steam Reforming Test Station

Both methanol and methane steam reforming trials take place in the steam reforming test

station illustrated in Figure 10. In the case of methanol steam reforming, a volumetric flask is filled

with a mixture of 1 mole of water and 1 mole of methanol. This mixture is pumped through the

system at a flow rate determined by the user using the control pump. For methane steam reforming,

the water flow rate is set through the control pump while the methane flow rate is set directly on

the LabVIEW interface.

The mixture of water and fuel, either methanol or methane, then passes through the mixer

vaporizer. This step in the process ensures that the water and the fuel are well mixed together

which is particularly important in the case of methane steam reforming where the reactants enter

the system separately. It also heats up and vaporizes the fuel and water mixture to ensure that they

enter the reformer as gases. The temperature of the mixer/vaporizer component is set at 250 °C for

methanol steam reforming or 175 °C for methane steam reforming.

The gas mixture, now at a higher temperature, travels through heated transfer lines to the

furnace, the temperature of these lines is set at 200 °C. This ensures that the mixture stays at high

temperature and avoids water condensation. The reactant mixture then passes through the reformer

and reacts overt the catalyst plate. The reformer is placed within an electric furnace which is set at

the desired reforming temperature. Figure 11 illustrates the reformer sitting in the furnace before

trials.

32

Finally, the product gas mixture exits the reformer and passes through a condenser which

is set at 1 °C. As the product gases flow through the condenser, excess steam is condensed and

accumulates at the bottom of the condenser. The dry product mixture then travels through a gas

chromatograph (Hewlett-Packard, USA) for composition analysis which is done through the Peak

simple software (SRI Instruments, USA). The gas chromatograph operating conditions are found

in Table 3.

Figure 10 – Schematic of the steam reforming test station.

33

Table 3 - Gas chromatograph operating conditions

Initial temperature 40 °C

Initial hold time 3 min

Ramp rate 8 °C /min

Final temperature 250 °C

Final hold time 5 min

Carrier gas 8.5 % H2/He

Carrier gas flow 10 cc/min

Sample size 1 mL

Figure 11 – Reformer installed in the furnace before steam reforming trials.

34

A variety of monitoring devices, thermocouples, pressure gauges and flow meters, are

found within this system. These are connected to a computer through a SCXI 1000 data acquisition

system (National Instruments, USA). They are monitored, and the data is collected, using a custom

LabVIEW (National Instruments, USA) interface which is pictured in Figure 12.

35

Figure 12 – Screen capture of the LabVIEW interface.

36

The FBGs are connected to a commercial optical interrogator, Micron Optics SM125

(Micron Optics, USA) and the data is collected through the ENLIGHT software (Micron Optics,

USA). The specifications of the SM125 are tabulated in Table 4.

Table 4 – SM125 specifications

Number of optical channels 4

Sampling frequency 2 Hz

Wavelength range 1510 - 1590 nm

Wavelength stability 1 pm

Wavelength accuracy 1 pm

Optical connectors FC/APC

Typical FBG sensor capacity 60-120

3.3. Experimental Procedure

3.3.1. Methanol Steam Reforming Trials

The methanol steam reforming trials are comprised of two steps. First, temperature

sensitivity of each FBG is determined. Since the methanol steam reforming trials are to take place

between 250 °C and 280 °C, a similar temperature range is used for characterization. The

characterization takes places in an electric furnace where the temperature is varied between 225

°C and 290 °C. The multiplexed FBGs and a Type K thermocouple are placed side by side in a

furnace. The temperature is first set to 225 °C and maintained until steady state is achieved. The

furnace temperature is then set to increase to 290 °C and once it has reached this desired

temperature, it is maintained until steady state is achieved. The furnace is then set to cool down to

225 °C. The thermocouple temperature is recorded every 0.5 s using LabVIEW while the Bragg

wavelength is recorded at the same rate but using the ENLIGHT software, these values are then

37

compared to obtain the temperature sensitivity of each FBG. This process is repeated five times

and the average temperature sensitivity is then used for the methanol steam reforming experiments.

Once the temperature sensitivity of the FBGs is determined, it is possible to use the sensors

in the methanol steam reformer. The thermocouples are first inserted in the holes at the bottom of

the reformer. The latter is then placed in the furnace and the fiber is inserted through the pre-made

hole. The reformer is then closed and the bolts are tightened. The furnace is brought up to the

desired reforming temperature, or furnace temperature, and left overnight to ensure that the entire

content of the furnace reaches a uniform temperature. During this time, a small flow, 0.010 ml/min,

of the water-methanol mixture is pumped through the system as to not damage the catalyst.

Once the temperature in the reformer has steadied over a long period of time, which will

be referred to as the stabilisation period, the flow rate is increased from 0.010 ml/min to the first

flow rate to be used for reforming. Once this is done, the product mass flow rate is monitored until

it is determined that the water-methanol mixture has reached the reformer and reforming is taking

place. The flow rate is further maintained at its current value until the two different gas samples

are analysed by the gas chromatograph. At this time, a new flow rate is entered in the control pump

and this flow rate is maintained until the reactants have reached the reformer and two different gas

samples are analyzed by the gas chromatograph. For every furnace temperature, a minimum of

three such water-methanol flow rates are selected for testing. Once experimental data has been

collected for a minimum of three flow rates, the furnace temperature is set to a new value and the

whole process is repeated to ensure that the temperature in the reformer is steady. This is done for

38

a total of three different furnace temperature, 250 °C, 265 °C and 280 °C. The test conditions are

summarized in Table 5 where time on-line represents the time since the temperature was first set

at 250 °C.

Table 5 – Summary of test conditions for methanol steam reforming trials

Furnace

temperature

[°C]

Cumulative time

on-line

[hr:min]

Methanol/water

flow rate

[ml/min]

250 17:28 0.050

250 19:07 0.070

250 21:14 0.090

250 22:50 0.110

250 24:13 0.130

250 44:42 0.010

265 46:34 0.010

265 47:51 0.050

265 49:05 0.070

265 65:15 0.010

265 66:32 0.090

265 69:29 0.110

280 89:06 0.010

280 90:36 0.130

280 92:04 0.260

280 93:45 0.320

39

3.3.2. Methane Steam Reforming Trials

As in the methanol steam reforming trials, the thermocouples are first inserted in the

reformer which is then placed in the furnace. The fiber is then inserted through the hole pictured

in Figure 5, the reformer is then closed and the bolts are tightened.

Given the high temperature needed to reform methane, the first step for the methane steam

reforming trials is regeneration of the FBG array. The desired regeneration temperature is 900 °C

and the reformer is heated up in steps. The thermocouples and FBGs are monitored to ensure that

the temperature inside the furnace has reached the desired point and has begun stabilizing before

it is increased to the following step. The furnace temperature is first set to 450 °C, then 650 °C,

800 °C and finally 910 °C which is slightly above the regeneration temperature of 900 °C. This is

done to account for the heat loss in the system. Once the furnace has reached temperatures close

to 910 °C, the gratings will erase and then regenerate. The temperature is maintained at 910 °C

until the regeneration comes to a stop, meaning there is no more change in the reflectivity of the

gratings.

Once regeneration is completed, the temperature sensitivity of each FBG is determined.

Since the methane steam reforming trials are to take place between 650 °C and 825 °C, a similar

temperature range is used for characterization. The characterization of the regenerated FBG array

is to be done in-situ once regeneration is completed. The temperature sensitivity of the FBGs can

be affected by the regeneration process and therefore characterization must be done after the

regeneration is completed. The fiber is to be left untouched in the reformer as to avoid damage.

40

The temperature of the reformer is varied between two set points as described earlier for the

methanol steam reforming FBG characterization. The six Type-K thermocouples located in the

reformer are monitored by the LabVIEW interface mentioned previously while the Bragg

wavelength of each FBG is recorded by the ENLIGHT software. These values are then compared

to obtain the temperature sensitivity of each FBG.

Like the methanol steam reforming trials, reforming is done at three different furnace

temperatures although an attempt was made to obtain further data at a fourth furnace temperature,

825 °C, and will be discussed later in the results. The three furnace temperatures are 650 °C, 750

°C and 775 °C and a minimum of 4 flow rates are selected for each furnace temperature, the test

conditions are summarized in Table 6 below where time on-line represents the time since the

temperature was first set at 650 °C for reforming. For each furnace temperature, the experimental

procedure is as follows: the furnace temperature is set and left to stabilize for a minimum of 15

hours, the pump controlling the water flow rate is first set at the desired flow rate and is given a

few minutes to travel through the system. The methane flow rate is then set to the appropriate value

based on the water flow rate and the desired steam to carbon ratio, S/C. The S/C ratio is a ratio of

the moles of steam to the mole of the hydrocarbons used in the experiments. The product mass

flow rate is monitored until it is determined that the water-methane mixture as reached the reformer

and reforming is taking place. The flow rate is further maintained at its current value until the two

different gas samples are analysed by the gas chromatograph. At this time, the flow rate is set to a

new value and is maintained until two different gas samples have been analyzed by the gas

chromatograph. Once the experimental data for minimum of 4 flow rates has been collected, the

41

furnace temperature is set to a new value and left to stabilize. During this time, a low flow rate of

5 ml/min of methane is circulated through the reformer as to avoid damage to the catalyst.

Table 6 – Summary of test conditions for methane steam reforming trials

Furnace temperature

[°C]

Cumulative time on-live [hr:min]

Methane flow rate

[ml/min]

S/C Water flow rate

[ml/min]

650 15:08 0 0 0

650 17:24 25 3 0.060

650 19:09 40 3 0.096

650 20:35 60 3 0.144

650 22:11 90 3 0.217

750 39:04 5 3 0.012

750 41:31 25 3 0.060

750 43:48 40 3 0.096

750 45:11 60 3 0.144

750 46:50 90 3 0.217

775 62:35 5 1.5 0.006

775 64:09 25 1.5 0.030

775 65:41 40 1.5 0.048

775 67:17 30 1.5 0.036

775 68:57 20 1.5 0.024

775 70:30 60 1.5 0.072

Chapter 4 and 5 will present the results obtained from the trials descried in the previous two

subsections.

3.4. Uncertainty

As seen in Table 4, the highest sampling frequency of the SM125 is 2 Hz. For the

thermocouples, the highest possible sampling frequency, which is determined by the settings on

42

the LABView program, is 0.5 Hz. For both types of sensors, the temperature data was recorded

using the highest possible sampling rate. The uncertainty in the measurements performed by both

FBGs and thermocouples is expressed in terms of standard deviation. This is possible since the

sampling rate was high and the measurements were taken over extended periods of time, see

Appendix C for more details on the data processing. Appendix D summarises the uncertainty

values for all the results presented in Chapter 4 and 5.

43

Chapter 4 - Temperature Monitoring in Methanol Steam Reformers

4.1. Overview

The objective of this thesis is to demonstrate the potential use of Type I FBGs to monitor

the temperature inside a steam reformer during both methanol and methane steam reforming.

Chapter 4 presents and discusses the results of the methanol steam reforming trials for which the

methodology was previously detailed in Section 3.3.1. This includes characterization of the FBG

array and the temperature profiles as a function of position in the reformer for various operating

conditions.

4.2. Characterization

Following the methodology described in Chapter 3, the temperature sensitivity of each of the

seven FBGs found on the array is obtained. The temperature range used for the methanol steam

reforming trials is 250 °C to 280 °C. Due to the increase in temperature sensitivity of the FBG as

the temperature increases, a temperature range similar to the one used during the trials is used for

characterization. The temperature range used for characterization is from 225 °C to 290 °C. A

figure presenting the Bragg wavelength of each FBG as a function of temperature for one of the

five instances where temperature was ramped up from 225 °C to 290 °C can be found in Appendix

B.

44

For this small temperature range, the relationship between Bragg wavelength and

temperature is assumed to be linear and follows the format in Equation (4.1) where λ0 is the initial

Bragg wavelength, T is the temperature in degree Celsius and KT is the temperature sensitivity of

the FBG.

𝜆𝐵 = 𝑇𝐾𝑇 + 𝜆0 (4.1)

Table 7 summarises the initial Bragg wavelength and temperature sensitivity values of all

seven FBGs obtained during the five rounds of characterization. The values for each of five

characterization runs can be found in Appendix B.

Table 7 – Average characterization values obtained from the five rounds of characterization for

FBG 1-7

FBG 1 FBG 2 FBG 3 FBG 4 FBG 5 FBG 6 FBG7

λ0 (nm) 1529.082 1534.067 1539.005 1544.023 1548.964 1554.001 1558.892

KT (pm/°C) 13.8 13.9 13.9 14.0 14.0 14.2 14.1

These values are used from here on out to convert the change in Bragg wavelength to the

temperature change in the reformer.

45

4.3. Temperature change in the reformer

4.3.1. Temperature profile for each furnace temperature

The figures presented in this section illustrate temperature change as a function of position

in the reformer. The dimensionless parameter x/L is used to represent sensor position. L is the

length of the catalyst plate while x is the distance from the leading edge of the reforming plate to

the FBG. Thus, x/L=0 corresponds to the leading edge of the catalyst plate located near the gas

inlet.

The temperature profile inside the reformer is presented with a relative temperature scale.

As mentioned previously in Section 3.3.1, the reformer and furnace are left to stabilise for an

extended period of time, ~15 hours. After this stabilisation period, it is assumed that the furnace

temperature has reached steady state. At this time, a baseline for the temperature, or Bragg

wavelength, of each FBG is determined and further changes in the Bragg wavelength are compared

to this baseline and then converted to temperature using the temperature sensitivity of the FBG.

More details on the data processing are presented in Appendix C.

Figure 13 illustrates the temperature change as a function of position in the reformer for a

furnace temperature of 250 °C. The flow rates used at this furnace temperature are 0.070, 0.090,

0.110 and 0.130 ml/min.

46


temperature of 250 °C.

In this graph, an initial drop in temperature is observed at the inlet for all flow rates. The

magnitude of this change varies between -0.4 and -0.6 °C for all flow rates. This indicates that near

the gas inlet, the temperature is 0.4 °C to 0.6 °C lower than the temperature at the same location

during the stabilisation period described previously. The largest temperature change is observed

at position 0.8 (FBG 6) and has a magnitude of 0.75 to 1.05 °C, depending on the flow rate. The

data for each flow rate follows the same trend. Temperature gradually decreases and the largest

temperature drop is found at the position of FBG 6 followed by a small increase in temperature

near the gas outlet.

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)

0.070 ml/min

0.090 ml/min

0.110 ml/min

0.130 ml/min

47

The graphs in Figure 14 and Figure 15 illustrate the temperature change as a function of

position in the reformer for furnace temperatures of 265 °C and 280 °C.



In Figure 14, there is drop in temperature at the inlet of the reformer compared to the

baseline temperature. Similar to the trends shown in Figure 13, temperature change for a furnace

temperature of 250 °C, the largest temperature change for each flow rate is observed at the location

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (X/L)

0.070 ml/min

0.090 ml/min

0.110 ml/min

48

of FBG 6. This temperature change is of -1.28 °C, -0.26 °C and -1.19 °C for the flow rates of 0.070

ml/min, 0.090 ml/min and 0.110 ml/min respectively. This is followed by a small increase in

temperature at the location of FBG 7 near the gas outlet. Importantly, the temperature change at

each position along the length of the catalyst plate for a flow rate of 0.090 ml/min is smaller than

the temperature change for the other two flow rates. For example, near the gas inlet, the

temperature change is -0.17 °C for a flow rate of 0.090 ml/min, -1.11 °C for a flow rate of 0.070

ml/min and -0.87 °C a flow rate of 0.110 ml/min.



-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)

0.130 ml/min

0.260 ml/min

0.320 ml/min

49

In Figure 15, a trend similar to the one in Figure 13, temperature change for a furnace

temperature of 250 °C, can be observed. The initial temperature change is of -0.93 °C, -1.48 °C

and -2.38 °C for the flow rates of 0.130 ml/min, 0.260 ml/min and 0.320 ml/min respectively.

Unlike in Figure 13 and Figure 14, temperature change in the reformer for furnace temperatures

of 250 °C and 265 °C respectively, the largest temperature change in this case is registered at the

location of FBG 7 which is nearest the gas outlet. The magnitude of this temperature change is of

-1.46 °C, -1.96 °C and -3 °C for the flow rates of 0.130 ml/min, 0.260 ml/min and 0.320 ml/min

respectively. Another trend that can be observed in Figure 15 is the near constant temperature

change for the first four FBG locations followed by a sharp decrease which is maintained until the

end of the reformer plate.

In summary, the graphs for the temperature change as a function of position in the reformer

for all three reforming temperatures show a similar trend. At the gas inlet (x/L=0), there is a drop

in temperature compared to the baseline temperature. As we progress along the length of the

catalyst plate, the temperature change increases with each successive data point. For methanol

steam reforming at the furnace temperatures of 250 °C and 260 °C, the largest temperature change

occurs at FBG 6 and is then followed by a slight recovery of the temperature change. For a furnace

temperature of 280 °C, a similar trend is observed but the temperature change increases until the

seventh FBG and the largest temperature change is registered at this location.

50

4.3.2. Comparison between the FBG and the thermocouple reference measurements

The previous section introduced the temperature change measurements made with the FBG

at three different furnace temperatures and detailed the observable trends within these graphs. The

current section compares the measurements obtained by the FBGs to the reference measurements

obtained with the thermocouples. The data processing for the thermocouples can also be found in

Appendix C. Note that, while seven FBGs were used to obtain data along the length of the catalyst

plate, only six thermocouples were used for the reference measurements due to limitations

associated with the thermocouple connector plate.

The graph in Figure 16 presents the data obtained from both thermocouples and FBGs for

a furnace temperature of 250 °C. The data for the FBGs is illustrated by the circles and the

thermocouples measurements are illustrated by the line symbol as seen in the graph’s legend.

51


temperature of 250 °C measured by FBGs and thermocouples.

Figure 16 allows the comparison between the FBG measurements and the thermocouple

measurements for a furnace temperature of 250 °C. From this figure, a few trends can be observed.

First, the trend observed in Figure 13 for the temperature change as a function of position measured

by the FBGs for a furnace temperature of 250 °C is also observed in Figure 16. The initial

temperature change is present and the temperature change increases with along the length of the

catalyst plate. Secondly, most of the temperature changes measured by the thermocouples are

larger than those measured by the FBG at the same positions. The graph in Figure 17 allows

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)

0.070 ml/min - FBG 0.090 ml/min - FBG


0.070 ml/min - Thermocouple 0.090 ml/min - Thermocouple


52

comparison between the FBG and thermocouple measurements for a furnace temperature of 265

°C.



Similar trends can be observed in the graphs of temperature change along the length of the

catalyst plate measured by both FBGs and thermocouples for the furnace temperatures of 250 °C,

Figure 16, and 265 °C, Figure 17. Once again, the increasing temperature change as a function of

position, until the sixth FBG (or fifth thermocouple), was measured by the thermocouples as well

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)

0.070 ml/min - FBG

0.090 ml/min - FBG

0.110 ml/min - FBG

0.070 ml/min - Thermocouple



53

as the FBGs. The temperature change measured by the thermocouples is always greater than that

measured by the FBG at the same location. Compared to the graph for the furnace temperature of

250 °C in Figure 16, there is not point where the temperature change measured by the thermocouple

is less than the one measured by the thermocouple for a given location. Finally, the graph in Figure

18 presents the data for both FBG and thermocouples but for a furnace temperature of 280 °C.



-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)


0.320 ml/min - FBG 0.130 ml/min - Thermocouple


54

The trends observed in Figure 18 are as observed in Figure 17, furnace temperature of 260

°C, and described previously. Mainly, the trends measured by the FBG and the thermocouples are

nearly identical and the temperature change measured by the reference thermocouples is always

greater than that recorded by the FBGs at a given location.

In summary, there is a high level of consistency between the FBG and thermocouples

measurements. The trends observed in the FBG measurements and described previously in Section

4.3.1 are also observed in the FBG and thermocouple comparative graphs for each of the three

furnace temperatures presented in Figure 16, Figure 17 and Figure 18. Another observation made

from these three figures is that the temperature changes measured by the thermocouple are

consistently slightly larger than those measured by the FBG at the same locations.

4.3.3. Effect of furnace temperature and conversion percentage on the temperature profile

This section presents the effect of reforming or furnace temperature on the temperature

profile for a given flow rate as measured by the FBGs. The temperature profile is again illustrated

by the temperature change, relative to a baseline. Moreover, this section introduces the methanol

conversion percentage, XCH3OH, and its effects on the temperature profile. A sample calculation for

the methanol conversation percentage is presented in Appendix E.

The graph in Figure 19 illustrates the temperature profile for a flow rate of 0.110 ml/min

at two different furnace temperatures, 250 °C and 265 °C. The methanol conversion percentage

55

measured for both these furnace temperature is found in Table 8. The methanol conversation for a

furnace temperature of 250 °C and a flow rate of 0.110 ml/min is 80.3% and 82.1% for a furnace

temperature of 265 °C and a flow rate of 0.110 ml/min.


ml/min.

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)

250 °C

265 °C

56


=0.110 ml/min

Furnace temperature XCH3OH

250 °C 80.3 %

265 °C 82.1%

Figure 19 shows the effect of the furnace temperature on the temperature profile for a

given flow rate. In this figure, the data for the flow rate of 0.110 ml/min is presented since it was

used during trials at both 250 °C and 265 °C. The principal observation that can be drawn from

this graph is that, when the furnace temperature is higher, the magnitude of the temperature change

at any location is greater than the magnitude of the temperature change at the same location at a

lower furnace temperature. The methanol conversion percentage does not change significantly, i.e.

only 1.8%, between these two furnace temperatures. There is no observable effect on the

temperature profile due to this change in methanol conversion.

Figure 20 presents the data for the flow rate of 0.130 ml/min at two furnace temperatures.

Table 9 presents the methanol conversion percentage for the two sets of operating conditions. For

this flow rate, the methanol conversion percentage is 75.8% for a furnace temperature of 250 °C

and 71.4% for a furnace temperature of 280 °C.

57


ml/min.


=0.130 ml/min

Furnace temperature XCH3OH

250 °C 75.8 %

280 °C 71.4%

The flow rate of 0.130 ml/min was tested for the furnace temperatures of 250 °C and 280

°C. Similarly to what is observed on the comparative graph for a flow rate of 0.110 ml/min, Figure

19, the temperature change is greater when the furnace temperature is higher. Unlike the graph for

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)

250 °C

280 °C

58

the flow rate of 0.110 ml/min, there is a noticeable effect on the temperature profile with the change

in methanol conversion percentage. For the higher furnace temperature of 280 °C, the methanol

conversion percentage is a few percent smaller than for the temperature of 250 °C. For 250 °C, the

temperature change is greater at the location of FBG 6 and recovers slightly near the gas outlet, at

FBG 7. This is not the case for the furnace temperature of 280 °C where the endothermic effect

has shifted slightly more toward the end of the catalyst plate and there is no return to lower

temperature changes near the gas outlet. The reason for this, and the other trends observed in the

figures of Section 4.3, is discussed in the following section.

In summary, the figures presented in Section 4.3.3 allow the following observation. When,

at a given flow rate, the reforming temperature increases, temperature change at all points in the

reformer increases. It is also possible to observe that when the methanol conversion percentage

diminishes the endothermic effect, position of the largest temperature drop, is moved closer to the

gas outlet.

4.4. Discussion

In the preceding sections, the results of the methanol steam reforming trials are presented. In

this section, these results are analyzed and discussed.

Methanol steam reforming is an endothermic reaction. Therefore, as the reactant mixture

encounters the catalyst at the leading edge of the catalyst plate, a zone of low temperature is

59

expected at this location. This is pictured in Figure 1 . The temperature inside the reformer is then

expected to return to its pre-reforming baseline as the flow of products moves along the length of

the catalyst without further reaction. The results shown previously for three different furnace

temperatures all show the largest temperature change in the second half of the reformer which is

contrary to this expectation.

Upon completion of the experimental work, the reformer was opened for examination. It was

discovered that the catalyst powder which is used for methanol steam reforming had been

displaced by the flow of reactants and had accumulated in the second half of the reformer, as shown

in Figure 21. Therefore, since this is where the catalyst powder is located, it is logical that the

largest temperature change would happen at some point in the second half or the reformer. While

this was unexpected, it demonstrates the ability of the FBG to measure accurately the temperature

changes inside the reformer.

Figure 21 - Photograph of the inside of the reformer after the methanol steam reforming trials.

60

A decrease in temperature near the gas inlet was observed in the graphs of temperature change

along the length of the reformer for all three furnace temperatures. Since there is no catalyst

powder in the first half of the reformer, a change in temperature at this location is unexpected. It

is believed that this temperature change in a location where there is no catalyst powder is caused

by the difference in temperature between the reactant stream and the reformer. The temperature

difference between the reactant stream and the reformer is caused by the temperature of the heated

transfer line in which the reactant stream flows, 200 °C, which is lower than the furnace

temperature.

The temperature change measured by the FBG at a furnace temperature of 265 °C and a flow

rate of 0.090 ml/min is quite different than the one measured for the flow rates of 0.070 ml/min

and 0.110 ml/min at the same furnace temperature. It is believed that the timing of the experiments

caused this. Looking back on Table 5, it can be seen that the testing for flow rate of 0.090 ml/min

was done following a long period of inactivity compared to the flow rate of 0.070 and 0.110

ml/min, which were done at the end of a day of experiment. It is possible that the conditions in the

reformer had not fully stabilised yet and that this could explain the lower temperature change for

the flow rate of 0.090 ml/min.

In Figure 15, temperature change along the length of the catalyst plate for a furnace temperature

of 280 °C, the temperature change increases at a given location when the flow rate is also increased

significantly. When there is an increase in flow rate, the amount of product entering the reformer

for a given period is increased as well. Since there is more reactant flowing over the catalyst plate,

61

the total amount of heat absorbed by the reaction increases which causes a larger drop in

temperature at this point in the reformer.

Since this work represents the first time FBGs are used to monitor temperature in a methanol

steam reforming, it is important to compare the data measured by the FBG to another type of

sensors. In this case, Type-K thermocouples, which were also during Aida Khosravi’s work using

the same reformer were used as reference measurements [75]. The comparison between these two

sets of measurements, which are performed simultaneously, are presented in Figure 16, Figure 17

and Figure 18 for the furnace temperatures of 250 °C, 265 °C and 280 °C respectively. The

temperature changes measured by both types of sensors are in agreement and confirm the validity

of the FBG measurements.

With few exceptions, the temperature change measured by the thermocouples is greater than

the temperature change measured by the FBGs. Since the thermocouples are embedded within the

base of the reformer, it is believed that they do not adjust as quickly as the FBG to the changes in

the room temperature. These changes in room temperature are addressed in Appendix C. This

could cause the baseline for the thermocouple measurements to be slightly higher. This could, in

turn, translate to greater temperature changes later on during the experiments.

When comparing the temperature change in the reformer for a given flow rate at two different

furnace temperatures, the principal observation is that, for the same flow rate, the temperature

change is greater when the furnace temperature is greater. This can be explained by the hydrogen

62

yield increase at higher temperature as mentioned in Chapter 1. A higher hydrogen yield implies

an increasing need for heat to power the endothermic chemical reaction which translates to a larger

temperature change in the reformer.

The uncertainty values for both the FBGs and thermocouple measurements are presented in

Appendix D. For the thermocouple measurements, the value of uncertainty is of 0.1 °C or less. The

uncertainty for the FBG measurements varies from 0.1 and 0.2 °C in the first half of the reformer

to 0.3 and 0.4 °C in the second half of the reformer. For the FBG measurements, it is possible to

see that the closer to the reaction site they are, the bigger the uncertainty of the measurements are.

The magnitude of the uncertainty in the second half of the reformer, 0.2-0.4 °C, is not significant

when compared to the temperature change at a given location. It is also important to note that the

standard deviation is calculated only for a one minute window and therefore only captures the

short term error. Some possible long term cause of experimental errors are also presents and worth

mentioning. These include: the accuracy of the flow pump, the variation in room temperature and

the movement of the catalyst powder toward the second half of the reformer.

In summary, the FBGs were able to measure the temperature change inside the reformer during

methanol steam reforming. When the catalyst powder was blown in the second half of the reformer

during trials, this was seen on the temperature change as a function of position in the reformer.

Moreover, the measurements done by the FBG illustrate the expected effect of increasing the

furnace temperature or the flow rate on the temperature change along the length of the catalyst

plate. Finally, the trend observed from both the thermocouples and the FBGs are in good

63

agreement. These results confirm the potential of using FOS based on FBGs for temperature

monitoring in methanol steam reforming.

64

Chapter 5 - Temperature Monitoring in Methane Steam Reformers

5.1. Overview

The objective of this thesis is to demonstrate the potential use of FOS based on FBGs to

monitor the temperature inside a steam reformer during both methanol and methane steam

reforming. Chapter 4 presented and discussed the results for the methanol steam reforming

experiments. Chapter 5 focuses on presenting and discussing the results of the methane steam

reforming trials for which the methodology was previously detailed in Section 3.3.2. This includes

regeneration of the FBGs, characterization of the FBG array and temperature profiles as a function

of position in the reformer for various operating conditions.

5.2. Regeneration

Methane steam reforming takes place at temperatures higher than the operating temperature of

Type I FBGs which were used in the methanol steam reforming trials. Due to the many advantages

of Type I FBGs over Type II FBGs, as described in Chapter 2, this type of FBG was also preferred

for the methane steam reforming experiments. The first step to successfully prepare the sensors

for the environment to which they will be subjected is to perform the regeneration of the FBGs.

The regeneration procedure used for the methane steam reforming experiments is described in

Chapter 3. Figure 22 illustrates the progress of the regeneration process through the power of the

grating and the temperature over time. The power of the grating is defined as the difference

between the peak of the Bragg wavelength and the noise floor. In this graph, the time zero

65

represents the time at which the regeneration process was begun, when the temperature was set to

increase to the first intermediate step as described in Chapter 3.

Figure 22 - Power and temperature over time during the FBG regeneration.

Figure 22 shows that the FBGs did not all regenerate simultaneously. Moreover, regeneration

began before the temperature in the reformer reached the target reforming temperature of 910 °C.

Figure 23 compares the full spectrum of this FBG array before and after regeneration.

66

Figure 23 - Full spectrum of the FBG array before and after regeneration.

The complete spectrum of the FBG array to be used for temperature monitoring in the

methane steam reformer is presented in Figure 23. The shift towards longer wavelengths of all

seven Bragg wavelengths is caused by the temperature increase in the furnace. Moreover, it is

possible to observe in Figure 22 and Figure 23 that the strength of the FBG is lower following

regeneration, as expected and discussed in Chapter 2.

67

5.3. Characterization

Before the regenerated FBGs are used to monitor the temperature change in the methane steam

reformer, the temperature sensitivity of these FBGs must be determined. The characterization

described in Section 3.3.2 was performed on another regenerated FBG array which was initially

supposed to be used in the methane steam reformer. As shown in Appendix F, complications arose

and it was not possible to use this array to monitor the temperature change in the methane steam

reformer.

The methane steam reforming experiments were performed at the test station located on

campus at Queen’s University. The reformer was available for a short time frame during which

the experiments are to be performed. Due to the complications arising from the first regeneration,

there was not enough time available to perform in-situ characterization of a second FBG array in

the reformer and obtain experimental data during methane steam reforming. Given the objective

of this thesis to prove the feasibility of using FBGs as temperature sensor in steam reforming, it

was deemed a priority to obtain experimental data during methane steam reforming and in-situ

characterization of the FBG array was not performed.

The temperature sensitivity of the regenerated FBG array to be used to monitor temperature in

the methane steam reformer is needed to convert the change in Bragg wavelength to a temperature

change. Since it is known that the regeneration may affect the temperature sensitivity value and

that the temperature sensitivity of an FBG will increase with increasing temperature, a different

68

characterization method was designed to obtain a close approximate of the temperature sensitivity

of the regenerated FBGs. First, a single FBG was characterized between 230 °C and 290 °C. This

was done in the same way as for the sensors used in methanol steam reforming and is described in

Section 3.3.1. The average temperature sensitivity of this FBG pre-regeneration was 13.5 pm/°C.

Moreover, pre-regeneration characterization of the FBG array to be used in the methane steam

reformer was performed between 230 °C and 290 °C. The average temperature sensitivity values

of the seven FBGs on the array are shown in Table 10. All values for the five pre-regeneration

characterization runs for both the single FBG and the FBG array are available in Appendix B.

Table 10 – Average pre-regeneration characterization values of the FBG array

FBG 1 FBG 2 FBG 3 FBG 4 FBG 5 FBG 6 FBG 7

Temperature

sensitivity (pm/°C)

13.9

13.9

13.9

14.1

14.2

14.2

14.2

While the pre-regeneration temperature sensitivity of the FBGs on the array are not identical,

these values range only from 13.9 pm/°C to 14.2 pm/°C. These values are slightly higher than the

temperature sensitivity of the single FBG which is 13.5 pm/°C.

Following pre-regeneration characterization, the single FBG was regenerated at high

temperature, as described in Chapter 3 and in Section 5.2. Following this regeneration, a new set

of characterization tests were performed to see how the temperature sensitivity was affected by the

69

regeneration. The temperature sensitivity of the single FBG was determined over three different

temperature range, 300 °C to 900 °C, 500 °C to 900 °C and 700 °C to 900 °C. The average

temperature sensitivity values of the single FBG over these three temperature ranges are shown in

Table 11. All values can be found in Appendix B.

Table 11 – Average temperature sensitivity of the single FBG for different temperature ranges

Temperature range 300 °C to 900 °C 500 °C to 700 °C 700 °C to 900 °C

Temperature

sensitivity (pm/°C)

15.2

15.6

16.0

The average temperature sensitivity of the single FBG varies between 15.2 pm/°C to 16.0

pm/°C based on the temperature range. These values indicate that for higher temperatures and

following regeneration, the temperature sensitivity of an FBG increases from ~13-14 pm/°C to

~15-16 pm/°C. Based on the pre-regeneration temperature sensitivity of the single FBG and the

FBG array as well as the post-regeneration temperature sensitivity of the single FBG, a temperature

sensitivity of 15 pm/°C for all FBGs found on the array was assumed to be a reasonable

approximation. This value is used to convert the change in Bragg wavelength during the methane

steam reforming experiments to change in temperature.

70

5.4. Temperature change in the reformer

5.4.1. Temperature drift

After the experiments, it was discovered that the temperature inside the reformer was

increasing during the methane steam reforming experiments despite the stabilization period

described in Section 4.3. Therefore, a drift correction factor was calculated individually for

each flow rate and is applied to all methane steam reforming data presented in Section 5.4.

The graph in Figure 24 illustrates the temperature drift as measured by FBG 1 during the

experiments at a furnace temperature of 650 °C. In this figure, a time of 0 hours represents the

time at which experiments were started for the first methane flow rate, 25 ml/min, at this

furnace temperature.

71

Figure 24- Wavelength over time of FBG 1 during experiments at 650 °C showing the

temperature drift inside the reformer

The goal of the drift correction factor is to determine the increase in temperature between

the baseline temperature and the time at which the data sample is taken. The drift correction factor

is then subtracted from the temperature change measured by the FBGs and the thermocouples. A

detailed description of how the correction factor is calculated and applied is presented in Appendix

G.

72

5.4.2. Temperature profile for each furnace temperature

The graphs for the temperature change in the reformer as a function of position during

methane steam reforming are presented in this section. As in Section 4.3, the dimensionless

variable x/L is used to represent position. L is the length of the catalyst plate while x is the distance

from the start of the reforming plate to the FBG. Thus, when x/L=0, it represents the leading edge

of the catalyst plate located near the gas inlet. The temperature profile inside the reformer is

presented with a relative temperature scale. The baseline for the temperature of the reformer is

obtained in the same manner as was described for the methanol steam reforming trials in Section

4.3. The data processing methods are presented in Appendix C.

The graph in Figure 25 represents the temperature change as a function of position in the

reformer for a furnace temperature of 650 °C. For this furnace temperature, the S/C ratio is of 3

and the methane flow rates used for testing are 25 ml/min, 40 ml/min, 60 ml/min and 90 ml/min.

73



Several trends are apparent in Figure 25. First, the largest temperature drops are measured by

either the second or third FBG. Beyond this point, the temperature change becomes positive. For

the flow rate of 25 ml/min, this temperature change hovers around 0 °C but as the experiments

progress, the temperature change near the gas outlet (FBG 7) reaches up to ~1 °C. For the flow

rate of 90 ml/min, there is a difference of ~1.8 °C between the lowest and highest temperature

change.

-4

-3

-2

-1

0

1

2

3

4

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)

25 ml/min

40 ml/min

60 ml/min

90 ml/min

74

Figure 26 illustrates the temperature change in the reformer for a furnace temperature of 750

°C. As for the furnace temperature of 650 °C, the S/C ratio used for the experiments performed at

750 °C is 3. The flow rates, 25 ml/min, 40 ml/min, 60 ml/min and 90 ml/min of methane, are

identical to those used for the experiments at 650 °C.



In Figure 26, the largest temperature drop is recorded by FBG 6 for the flow rates of 40 ml/min,

60 ml/min and 90 ml/min. In the case where the methane flow rate was 25 ml/min, the largest

-4

-3

-2

-1

0

1

2

3

4

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)

25 ml/min

40 ml/min

60 ml/min

90 ml/min

75

temperature drop is measured by FBG 4 although, for this flow rate, all temperatures hover slightly

above or below 0 °C and no clear zone of high or low temperature is observed. For a methane flow

rate of 90 ml/min, the difference between the largest temperature drop and highest temperature

change, recorded by FBG 1, is ~ 6 °C. This is more than three times the difference measured during

the experiments performed at a furnace temperature of 650 °C.

In Figure 27, the temperature change as a function of position for the furnace temperature of

775 °C is presented. For this reforming temperature, the S/C ratio used is 1.5. To facilitate

comparison with the data obtained for reforming temperatures of 650 °C and 750 °C, the same

methane flow rates are used, with the exception of 90 ml/min, and the steam flow rate is adjusted

according to the S/C ratio. Two additional methane flow rates are tested for the reforming

temperature of 775 °C. These are 20 ml/min and 30 ml/min. Finally, the legend in Figure 27 lists

the flow rates in increasing order for clarity reasons. It is important to note that these are not the

order in which they were tested. This order, as presented in Table 6, is 25 ml/min, 40 ml/min, 30

ml/min, 20 ml/min and finally 60 ml/min.

76



The trends observed in Figure 27 are similar to the ones described earlier for Figure 25 for the

furnace temperature of 650 °C. The largest temperature drop is located near the gas inlet and

measured by FBG 2. The magnitude of this temperature drop is larger, up to ~1.5 °C, than the one

measured during the experiments done with a reforming temperature of 650 °C. In Figure 27, the

difference between the largest temperature drop and the highest temperature change is of ~3 °C,

which is also more significant than the value, ~1.8 °C, measured for the reforming temperature of

650 °C. After the initial temperature drop, the temperature change in the reformer approaches zero

-4

-3

-2

-1

0

1

2

3

4

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)

20 ml/min

25 ml/min

30 ml/min

40 ml/min

60 ml/min

77

around the mid-point of the reformer. Following this, the temperature change is positive and

reaches up to ~1.8 °C for the flow rate of 40 ml/min. Depending on the flow rate, the highest

temperature change is either measured by FBG 1, FBG 6 or FBG 7. Finally, it is possible to observe

that despite small differences, the trend measured for the temperature change as a function of

position in Figure 25 and Figure 27 are quite similar and differ greatly from Figure 26.

In summary, the graphs of the temperature change as a function of position in the reformer for

reforming temperatures of 650 °C and 775 °C presented in Figure 25 and Figure 27 respectively

show the same trends. The largest temperature drop is registered near the gas inlet. Following this,

the temperature change becomes positive and reaches its maximum near the gas outlet.

The results obtained for the reforming temperature of 750 °C show a different temperature

change profile along the length of the catalyst plate. In this case, the highest temperature change

is measured by the first FBG, at the gas inlet, and the largest temperature drop is registered by an

FBG located near the gas outlet. The graph for the reforming temperature of 750 °C, Figure 26,

shows a trend opposite to the trend observed for the reforming temperature of 650 °C and 775 °C.

5.4.3. Comparison between the FBG and the thermocouple reference measurements

In Section 5.4.1, the graphs for the temperature changes as a function of position in the

reformer were presented for all three furnace temperatures, 650 °C, 750 °C, 775 °C. As for the

78

methanol steam reforming experiments, thermocouples are embedded in the base of the reformer

and used as reference measurements. This section will compare the measurements obtained with

the FBGs and the reference measurement performed with the thermocouples. The data processing

for the thermocouple is explained in detail in Appendix C. As for the methanol steam reforming

experiments, only six thermocouples are used during the methane steam reforming experiments

due to limitations associated with the thermocouple connector plate. Therefore, there is no

thermocouple located below FBG 2.

Figure 28 compares the FBG and thermocouple measurements for a furnace temperature

of 650 °C. The FBG data points are represented by circle while the thermocouple measurements

are represented by a line as can be seen in the legend of the graph.

79



In Figure 28, the thermocouple data follows a trend similar to the FBG trend described in

Section 5.4.1. Depending on the flow rate, the largest temperature drop is captured by one of the

first two thermocouples, near the gas inlet. The temperature change then hovers near zero at the

location of the third thermocouple (FBG 4) which is near the mid-point of the reformer. In the

second half of the reformer, the temperature change is positive and reaches a maximum value near

the gas outlet, thermocouple 6. The only exception to this is the measurement for the flow rate of

25 ml/min which reached its maximum at the location of thermocouple 5. The magnitude of the

-4

-3

-2

-1

0

1

2

3

4

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)

25 ml/min - FBG 40 ml/min - FBG


25 ml/min - Thermocouple 40 ml/min - Thermocouple


80

temperature change measured by the thermocouple is similar to the magnitude of the

measurements obtained from the FBG and this observation is valid for all four flow rates.

Figure 29 compares the FBG and thermocouple measurements for a furnace temperature of

750 °C.



-4

-3

-2

-1

0

1

2

3

4

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)





81

The measurements taken with the thermocouples when the furnace temperature is set to 750

°C can be seen in Figure 29. Unlike Figure 28, which compares the thermocouple and FBG

measurements for a furnace temperature of 650 °C, the temperature changes measured by the

thermocouples follow a different trend than the data taken with the FBGs. The largest temperature

drop measured by the thermocouple is located near the gas inlet, thermocouple 1. From there, the

temperature change increases in a constant manner until it reaches its highest value near the gas

outlet, thermocouple 6. The magnitude of the temperature changes measured by the thermocouple

is much smaller, ranging from -0.5 °C to 0.5 °C, than the temperature change measured by the

FBGs.

Finally, the graph in Figure 30 compares the measurements obtained with the FBG and

thermocouples for a furnace temperature of 775 °C.

82



The observation made from Figure 28, comparison between thermocouples and FBGs for a

furnace temperature of 650 °C, are similar to the trends observed in Figure 30. In this latest figure,

the data obtained through the thermocouple measurements does not follow the exact trend of the

data measured by the FBGs but a similar one. The largest temperature drop occurs at the location

of thermocouple 1, near the gas inlet, while the largest temperature drop measured by an FBG

occurs at the position of FBG 2. The temperature change measured by the thermocouple increases

constantly along the length of the catalyst plate. The highest temperature change is registered near

-4

-3

-2

-1

0

1

2

3

4

0 0.2 0.4 0.6 0.8 1

Tem

per

atu

re c

han

ge (

°C)

Position (x/L)



60 ml/min - FBG 25 ml/min - Thermocouple



83

the gas outlet by thermocouple 6. Like the data collected for a furnace temperature of 750 °C, the

magnitude of the measurements obtained by the thermocouples, ranging from -0.3 °C to 0.2 °C, is

much smaller than the magnitude of the measurements obtained with the FBGs.

In summary, the reference measurements obtained with the six thermocouples located along

the length of the catalyst plate show a similar trend for the furnace temperatures of 650 °C and 775

°C. Despite the similar trend, the magnitude of the temperature change measured by the

thermocouples is much smaller than that measured by the FBGs. Finally, for one of the three

furnace temperatures, 750 °C, the trend observed from the thermocouple measurements is different

than the one obtained with the FBGs.

5.4.4. Effect of furnace temperature and S/C ratio on the temperature profile

During the methane steam reforming experiments, three methane flow rates, 25 ml/min, 40

ml/min and 60 ml/min, were tested for each furnace temperature and the chosen S/C ratio. This

allows for comparative graphs. During the methane steam reforming trials, three operating

conditions were varied. These are furnace temperature, flow rate and steam to carbon ratio, S/C

ratio.

The graphs in Figure 31 illustrate the temperature changes in the reformer for two different

furnace temperatures, 650 °C and 775 °C, and two different S/C ratios. For a furnace temperature

of 650 °C, the S/C of choice was 3. For a furnace temperature of 775 °C, a S/C of 1.5 was selected.

84

Figure 31(a) presents this graph for a flow rate of 40 ml/min and Figure 31(b) for a flow rate of 60

ml/min.

Figure 31 - Temperature change in the reformer as a function of positions for two different

furnace temperatures and S/C ratio (a)- flow rate of 40 ml/min (b)- flow rate of 60 ml/min.

The graph in Figure 31(a) illustrates the temperature change as a function of position in the

reformer for two different sets of operating conditions when the flow rate is set to 40 ml/min.

These are a reforming temperature of 650 °C and a S/C ratio of 3 and secondly, a reforming

temperature of 775 °C and a S/C ratio of 1.5. The trend observed for both sets of operating

conditions is quite similar. The largest temperature drop is registered near the gas inlet, FBG 2,

and the temperature change reaches 0 near the mid-point of the reformer before increasing and

registering the highest temperature change near the gas outlet, FBG 7. The one noticeable

85

difference between the two data sets is that the magnitude of the temperature change for the

reforming temperature of 775 °C and the S/C ratio of 1.5 is consistently higher than the values

obtained for the reforming temperature of 650 °C and the S/C of 3. The graph in Figure 31(b) also

illustrates the temperature change as a function of position in the reformer for the same operating

conditions but for a flow rate of 60 ml/min. The same observations that are detailed above for

Figure 31(a) can be made from Figure 31(b). There is one exception and it is the temperature

change recorded by FBG 1 for the reforming temperature of 775 °C and the S/C ratio of 1.5 which

is positive whereas the temperature change measured for temperature of 650 °C and the S/C of 3

is a temperature drop, negative value. Moreover, the magnitude of this data point for the reforming

temperature of 775 °C and the S/C ratio of 1.5 is of 1.7 °C which is several times higher than the

temperature change of -0.2 °C registered for the reforming temperature of 650 °C and the S/C of

3.

In summary, when comparing the temperature change as a function of position in the reformer

for two varying sets of operating conditions, reforming temperature and S/C ratio, the temperature

change profile along the length of the catalyst plate is nearly identical for both operating

conditions. The magnitude of the temperature change for the higher reforming temperature, 750

°C, and lower S/C ratio, 1.5, is more important the one for the lower reforming temperature, 650

°C, and higher S/C ratio, 3.

86

5.4.5. Secondary erasure

Following the experiments performed at a reforming temperature of 775 °C, the furnace

temperature was set to increase to 825 °C with the objective to perform additional testing at this

temperature. As is described in Section 4.3 for the methanol steam reforming experiments, the

furnace and reformer where then left to stabilise for a period of ~15 hours. During this time, the

Bragg wavelength of each FBG along with the full spectrum of the array were monitored. After

the stabilisation period, it was discovered that all seven FBGs on the array had undergone

secondary erasure. The graph in Figure 32 illustrates this phenomenon through the full spectrum

of the sensor array at different points in time. In this figure, the time of zero represents the moment

at which the furnace is set to increase from 775 °C to 825 °C.

87

Figure 32 - Full spectrum of the FBG array undergoing secondary erasure while the furnace

temperature is increasing from 775 °C to 825 °C.

In Figure 32, it is possible to observe the full spectrum of the FBG array at the moment where

the furnace temperature is set to 825 °C. As the temperature inside the furnace increases from 775

°C to 825 °C, the Bragg wavelength shifts to higher values and this is observed in Figure 32. All

seven FBGs undergo simultaneous secondary erasure over a period of over 25 minutes. Since all

FBGs on the array used for testing had undergone secondary erasure, it was impossible to measure

the temperature change in the reformer during methane steam reforming at a temperature of 825

°C and therefore, experiments were stopped at this time.

88

5.5. Discussion

As the temperature in a methane steam reformer is above the operating temperature of a Type

I FBG, it is necessary to regenerate the FBGs before going through with the methane steam

reforming experiments. The FBG’s power and temperature over time for the regeneration of the

FBG are shown in Figure 22. The FBGs started erasing and regenerating before the final

regenerating temperature was reached, which took over 7 hours. Due to its size and the material

from which it is built, the reformer has a high thermal mass. Therefore, heating up the reformer to

the regeneration temperature takes a few hours. When regenerating an FBG on its own in a high-

temperature furnace, the regenerating temperature can be reached much faster, sometimes in one

to two hours depending on the regeneration procedure used.

Another thing to note from the power and temperature over time graph of Figure 22 is that not

all FBG regenerate simultaneously. This was unexpected and was not reported by Laffont et al. in

their work on regenerating multiplexed FBGs [68]. In their work, all FBGs regenerated

simultaneously once the desired regenerating temperature was reached. This temperature was

reached in approximately two and a half hours which is approximately one-third of the time taken

to reach the regenerating temperature in the experiments reported here.

The trend observed in the graphs for the reforming temperatures of 650 °C and 775 °C, Figure

25 and Figure 27, are nearly identical. As mentioned in Section 4.4, when the stream of reactants

first encounters the catalyst, this is where most of the reaction occurs. Methane steam reforming

is an endothermic reaction and therefore it is expected that the FBGs near the gas inlet would

89

measure an important temperature drop as the reaction absorbs the surrounding heat. After this

initial temperature drop, the temperature is expected to return to, or near, the baseline measured

before the experiments were started as the stream of products flow over the catalyst plate to the

gas outlet without further reaction. Unexpectedly, the temperature change measured by the FBG

in the second half of the reformer increases to positive values and reaches a maximum near the gas

outlet, measured by FBG 6 or 7. If the temperature in the reformer was at steady-state prior to the

experiments, such an increase in temperatures in the reformer would mean that heat is released

within the reformer. This is synonym of an exothermic reaction. Since it is known that methane

steam reforming is an endothermic reaction, this is not possible.

Another explanation for the positive temperature change in the second part of the reformer is

that the drift correction factor does not accurately represent the temperature drift in the reformer

throughout the experiments. The drift correction factor introduced to mitigate the effect of the

temperature drift during the experiments relies on the assumption that the whole reformer drifts

uniformly to high temperatures. This might not be the case given that the extremities of the

reformer are located closer to the edge of the furnace where heat loss is expected to be higher.

Despite the positive values for temperature change in the reformer, the FBGs captured the expected

temperature change profile along the length of the catalyst plate.

Since temperature monitoring in methane steam reforming is a new application for FBGs, the

measurements done by the FBGs are compared to the thermocouple measurements. This is

presented in Figure 28, Figure 29 and Figure 30 for the furnace temperatures of 650 °C, 750 °C

90

and 775 °C respectively. The thermocouple measurements for all three furnace temperatures show

the same trend. As mentioned previously, a positive temperature change in the reformer is

unexpected due to the endothermic nature of the reaction. This result is most likely caused by the

temperature drift in the reformer during the experiments and the limitations of the drift correction

factor. The trend captured by the thermocouple measurements confirms the zone of low

temperature near the gas inlet measured by the FBGs which corresponds to the area where the

endothermic steam reforming reaction takes place.

One major difference between the FBG and thermocouple measurements is the magnitude of

the temperature change. To help understand this, the maximum and minimum values of

temperature change, regardless of flow rate, measured by both FBGs and thermocouples at a

furnace temperature are summarized in Table 12.

Table 12 - Maximum and minimum temperature change recorded by both FBGs and

thermocouples regardless of flow rates

Reforming

Temperature

650 °C 750 °C 775 °C

Maximum Minimum Maximum Minimum Maximum Minimum

FBG 0.9 °C -0.7 °C 3.0 °C -2.8 °C 1.8 °C -1.5 °C

Thermocouple 0.7 °C -0.7 °C 0.6 °C -0.5 °C 0.2 °C -0.3 °C

91

The discrepancies between the magnitude of the measurements obtained from the

thermocouples and the FBGs for the furnace temperature of 750 °C and 775 °C may be due to by

the location of the thermocouples. The FBG are located directly under the catalyst plate and are

therefore very close to the reaction. As seen in Chapter 3, the thermocouples are embedded in the

base of the reformer and are separated from the FBG and the reaction site by a copper plug used

to seal the reformer. The thermal conductivity of copper is high, 400 W/m K, compared to stainless

steel 310, 18.7 W/m K at 500 °C. Therefore, the steel body of the reformer surrounding the cooper

plug would influence its temperature. The tip of the thermocouple is located just below the copper

plug and since the temperature of the copper plug reflects the temperature of the steel body of the

reformer, the thermocouples would measure this temperature and not accurately reflect than the

temperature at the reaction site.

In Figure 31, data for the reforming temperature of 650 °C and the S/C ratio of 3 is compared

to the data obtained from the experiments when the reforming temperature is 775 °C and the S/C

ratio is 1.5. In these graphs, it is possible to observe that the temperature drop near the gas inlet is

larger when the reforming temperature is increased from 650 °C to 775 °C. As mentioned in

Section 4.4, when the reforming temperature is increased for a given flow rate, the hydrogen yield

is increased as well. An increase in hydrogen yield necessitates more heat from the reaction. So

therefore, when there is an increased in the reforming temperature for a given flow rate, the

temperature change in the reformer should be greater. This is observed in Figure 31. Another

important thing to discuss for the graphs of Figure 31 is the change in S/C ratio between the two

data sets. During the reaction to produce hydrogen and carbon dioxide, as in Equation (1.1), the

92

stochiometric S/C ratio of methane steam reforming is 2.0. When producing hydrogen and carbon

monoxide, the S/C ratio, as seen in Equation (5.1), becomes 1.0 [77].

𝐶𝐻4 + 𝐻2𝑂 → 𝐶𝑂 + 3𝐻2 ∆𝐻2980 = 206 𝑘𝐽/𝑚𝑜𝑙 (5.1)

In commercial applications, the S/C ratio used during methane steam reforming would always

be greater than 2.0 to avoid the formation of carbon dioxide. For the purpose of this work, reducing

the S/C ratio below 2.0 was of interest. Indeed, the chemical reaction for steam reforming which

produces carbon monoxide and hydrogen is more endothermic than the reaction which produces

carbon dioxide and hydrogen. A higher endothermic value would create a larger temperature drop

in the reformer as the methane and the steam react over the catalyst near the gas inlet. The

combined effect of the higher reforming temperature and the lower S/C ratio creates a larger

temperature drop near the gas inlet. This is captured by the thermocouple as seen in Figure 31.

The uncertainty values for both the FBGs and thermocouple measurements are presented in

Appendix D. These uncertainty values are based on the standard deviation of the data during the

one minute period of data collection, as explained in Appendix D. Since the standard deviation is

calculated only for a one minute window, it only captures the short term error. Some possible long

term cause of experimental errors are also presents and worth mentioning. These include: the

accuracy of the flow pump and the methane flow rate controls, the variation in room temperature

and the drift in temperature observed during the trials.

93

For the thermocouple measurements, the value of uncertainty is of 0.1 °C or less. The

uncertainty for the FBG measurements varies from 0.1 and 0.2 °C for most of the measurements.

The magnitude of the FBG’s uncertainty in the reformer, 0.1-0.2 °C, is not significant when

compared to the temperature change at a given location. For the furnace temperature of 750 and

775 °C, the uncertainty increases for FBG 7 to values of up to 0.9 °C which is quite significant

when compared to the temperature measured by this FBG. It can be noted that this increase in

uncertainty in FBG 7 corresponds to the variation, and subsequent permanent increase, in the noise

floor level discussed in the following section. It is believed that the variation in noise floor level

affected the accuracy of the measurements performed by FBG 7.

Despite some issues arising during the experiments at a furnace temperature of 750 °C, the

results for the furnace temperatures of 650 °C and 775 °C presented and discussed in Chapter 5

demonstrate the ability of FBGs to monitor temperature changes in methane steam reformers. For

these two furnace temperatures, the FBGs were able to capture the profile of the temperature

change along the length of the catalyst plate for different operating conditions. The analysis of the

results presented thus far points to the feasibility of using FBG to monitor temperature in methane

steam reformers. A few issues arose during the methane steam reforming experiments and these

are discussed in Section 5.5.1.

94

5.5.1. Outstanding Issues

As mentioned in Chapter 2, regenerated FBGs have been shown to be stable for up to 2000

hours but the FBG array used in this work underwent secondary erasure after less than 100 hours

at temperatures above 600 °C [47]. Secondary erasure of regenerated FBG is not reported in the

literature. The cause of the secondary erasure has not been determined, however, it may be that the

time taken to bring the reformer, and thus the FBG array, to the desired regenerating temperature

played a role in the secondary erasure of the FBGs.

While the data for the reforming temperatures of 650 °C and 775 °C matches the expected

temperature profile along the length of the catalyst plate, the data for 750 °C looks quite different.

The profile of temperature change along the length of the catalyst plate does not match the expected

behaviour for an endothermic reaction. The temperature change along the length of the catalyst

plate shows an unlikely behaviour for an endothermic chemical reaction and goes against the trend

observed so far in this thesis.

During the experiments at a reforming temperature of 750 °C, it was observed that the noise

floor varied between two levels at random time intervals. Figure 33 presents the full spectrum of

the FBG array taken at two different times, 11 minutes apart. During the experiments, the noise

floor goes from the lower level to the higher level, and vice versa, instantaneously while the peak

of the Bragg wavelength remains at the same level.

95

Figure 33 - Full spectrum of the FBG array showing the change in the noise floor level.

It is important to note that during regeneration or erasure, the noise floor stays at a constant

level while the reflectivity of the grating decreases until the peak of the Bragg wavelength is

indistinguishable from the noise floor. Therefore, Figure 33 does not illustrate erasing of the FBGs

but rather some other issue which could have causes the erroneous measurements taken at a furnace


96

As described here, some issues arose while using regenerated FBGs to monitor the temperature

in the methane steam reformers. The cause of these issues is unknown and further work will be

needed. Despite these issues, two out of the three sets of results, furnace temperature of 650 °C

and 750 °C, presented in Chapter 5 demonstrate the viability of temperature sensors based on FBGs

in methane steam reformers. Chapter 6 will conclude this thesis and offer recommendations to

improve the work presented in thesis if further work on the subject is to be pursued.

97

Chapter 6 - Conclusions and Future Work

6.1. Conclusions

The use of fiber optics sensors such as FBGs to monitor the temperature in methanol and

methane steam reforming could provide important information about zonea of low temperature in

the reformers. The objective of this thesis is to demonstrate the feasibility of using FBGs to monitor

the temperature in steam reformers during both methanol and methane steam reforming.

An FBG array consisting of 7 FBGs placed at strategic positions along the length of the catalyst

plate was inserted in a modified metal-plate test reformer. Methanol steam reforming experiments

were performed for a total of 12 flow rates at three different reforming temperatures, 250 °C, 265

°C and 280 °C. The temperature during the reforming experiments was compared to a baseline

temperature at each position to obtain the profile of temperature change along the length of the

reformer. The FBG measurements were compared to the thermocouple measurements taken

simultaneously. The profile for the temperature change along the catalyst plate measured by the

FBG, alongside of a photograph of the catalyst powder inside the reformer shown in Figure 21,

show a zone of low temperature when the stream of reactant encounters the catalyst powder. This

is measured not only by the FBGs but by the thermocouples as well. This is also similar to the

modelling work done on the subject which shows a zone of low temperature at the inlet of the

reforming chamber [24]. Moreover, it was demonstrated that the FBGs were able to measure small

changes within the reformer. These changes are an increased in temperature drop when the furnace

temperature is increased and an increase in temperature drop when the flow rate is significantly

98

increased. The FBGs were also able to capture the largest temperature drop shift toward the end

of the reformer when the methanol conversion percentage was decreased.

To survive the operating temperature of methane steam reforming, the FBG array used for

temperature monitoring had to be regenerated. Regeneration was performed while the fiber was in

place in the reformer to avoid subsequent manipulation of the bare fiber. Regeneration in this

environment took over seven hours, which is much longer than the regeneration of FBGs done

without the reformer present. The regenerated FBG array was then used to monitor temperature in

the reformer during methane steam reforming experiments. A total of 13 sets of operating

conditions were tested. This includes three different reforming temperatures, 650 °C, 750 °C and

775 °C, and two different S/C ratios, 1.5 and 3. The results from these experiments, which include

a drift correction factor to account for the temperature drift in the reformer as the experiments were

progressing, demonstrate the possibility of using FBGs to monitor temperature in methane steam

reformers. Through the graphs of temperature change as a function of position along the length of

the catalyst plate and the comparison with temperature change data obtained from the

thermocouples, it is possible to conclude that FBGs were successful in capturing the area of low

temperature associated with the reaction zone near the gas inlet. Moreover, the FBGs were able to

capture an increase in the temperature drop as the reforming temperature was increased and the

S/C ratio was decreased below the stoichiometric value for the production of hydrogen and carbon

dioxide.

99

Despite the success in demonstrating the possibility to measure temperature in both

methanol and methane steam reforming, some issues were discovered. An attempt was made to

capture temperature data while reforming methane at a temperature of 825 °C. This results in the

FBG array going through secondary erasure during the stabilisation period. The instability of the

FBGs due to issues with the regeneration were also most likely the cause of the erroneous reading

obtained for the reforming temperature of 750 °C.

6.2. Recommendation for future work

The experimental work presented in this thesis introduces a new application for FBGs. To

further improve the temperature monitoring in methane steam reformers during methanol and

methane steam reformers, a few recommendations are presented in this section.

The zone of low temperature in steam reformers is located where the stream of reactant first

encounters the catalyst since this is where most the reaction take place. For this reason, the spatial

resolution of FBGs in the first third of the reformer was higher than in the rest of the reformer. To

gain further insight into this important section of the reformer, it would be possible to increase the

spatial resolution of the FBGs more by reducing the length of the FBG to 1 mm and decreasing

the spacing to the minimum allowed by the manufacturer.

Secondly, the use of a well-designed packaging system could serve a few purposes. First, the

FBG array could be regenerated in a stand-alone furnace which could significantly reduce the time

100

needed to bring the fiber to the regeneration temperature. Since regeneration renders the fiber

extremely brittle, any handling of the regenerated bare fiber would result in its shattering. The

packaging would eliminate the need to handle the bare fiber after regeneration and allow it to be

safely placed in the reformer. Regenerating the fiber before it is placed in the reformer has the

potential to eliminate the secondary erasure of the FBGs. If the FBG array is regenerated in its

packaging, it would also allow the FBGs to be characterized after regeneration but before being

placed in the reformer. Finally, and this is valid for both methanol and methane steam reforming,

a sealed packaging could allow the fiber to be placed directly on the catalyst plate rather than

below it. The fiber is currently placed under the catalyst plate to be protected from hydrogen but a

sealed packaging would eliminate the need for the fiber to be separated from the gases in the

reformer.

101

References

[1] IPCC, “Renewable Energy Sources and Climate Change Mitigation,” 2012.

[2] R. L. Fares and M. E. Webber, “The impacts of storing solar energy in the home to reduce

reliance on the utility,” Nat. Energy, vol. 2, no. 2, 2017.

[3] A. Bilodeau and K. Agbossou, “Control analysis of renewable energy system with

hydrogen storage for residential applications,” J. Power Sources, vol. 162, pp. 757–764,

2006.

[4] K. Agbossou, R. Chahine, J. Hamelin, F. Laurencelle, A. Anouar, J.-M. St-Arnaud, and T.

K. Bose, “Renewable energy systems based on hydrogen for remote applications,” J.

Power Sources, vol. 96, no. 168–172, 2001.

[5] S. R. Vosen and J. O. Keller, “Hybrid energy storage systems for stand-alone electric

power systems: optimization of system performance and cost through control strategies,”

Int. J. Hydrogen Energy, vol. 24, pp. 1139–1156, 1999.

[6] United States Environmental Protection Agency, “Global Greenhouse Gas Emissions Data

| Greenhouse Gas (GHG) Emissions | US EPA.” [Online]. Available:

https://www.epa.gov/ghgemissions/global-greenhouse-gas-emissions-data. [Accessed: 16-

Mar-2017].

[7] Government of Canada, “Environment and Climate Change Canada - Environmental

Indicators - Greenhouse Gas Emissions by Economic Sector.” [Online]. Available:

https://www.ec.gc.ca/indicateurs-indicators/default.asp?lang=en&n=F60DB708-1.

[Accessed: 16-Mar-2017].

102

[8] M. Ball and M. Wietschel, “The future of hydrogen – opportunities and challenges 5,” Int.

J. Hydrogen Energy, vol. 34, no. 2, pp. 615–627, 2009.

[9] P. Maniatopoulos, J. Andrews, and B. Shabani, “Towards a sustainable strategy for road

transportation in Australia: The potential contribution of hydrogen,” Renew. Sustain.

Energy Rev., vol. 52, pp. 24–34, 2015.

[10] N. Balta-Ozkan and E. Baldwin, “Spatial development of hydrogen economy in a low-

carbon UK energy system,” Int. J. Hydrogen Energy, vol. 38, no. 3, pp. 1209–1224, 2013.

[11] P. P. Edwards, V. L. Kuznetsov, W. I. F. David, and N. P. Brandon, “Hydrogen and fuel

cells: Towards a sustainable energy future,” Energy Policy, vol. 36, no. 12, pp. 4356–

4362, 2008.

[12] J. D. Holladay, J. Hu, D. L. King, and Y. Wang, “An overview of hydrogen production

technologies,” Catal. Today, vol. 139, pp. 244–260, 2009.

[13] Office of Energy Efficiency & Renewable Energy, “Hydrogen Resources | Department of

Energy.” [Online]. Available: https://energy.gov/eere/fuelcells/hydrogen-resources.


[14] Department of energy, “Multi-Year Research, Development and Demonstration Plan,”

2015.

[15] Uniongas, “Chemical Composition of Natural Gas.” [Online]. Available:

https://www.uniongas.com/about-us/about-natural-gas/Chemical-Composition-of-Natural-

Gas. [Accessed: 16-Mar-2017].

[16] National Energy Board, “NEB - Natural Gas.” [Online]. Available: https://www.neb-

103

one.gc.ca/nrg/sttstc/ntrlgs/index-eng.html. [Accessed: 05-Jul-2017].

[17] International Energy Agency, “Hydrogen Production and Distribution,” 2007.

[18] Canadian Gas Association, “Gas Stats.” [Online]. Available: http://www.cga.ca/gas-stats/.


[19] D. R. Palo, R. A. Dagle, and J. D. Holladay, “Methanol steam reforming for hydrogen

production,” Chem. Rev. 107 3992-4021, vol. 107, no. PNNL-SA-53246, pp. 3992–4021,

2007.

[20] T. L. De Souza, C. D. C. R. D. S. Rossi, C. G. Alonso, R. Guirardello, V. F. Cabral, N. R.

C. Fernandes-Machado, S. Specchia, M. S. Zabaloy, and L. Cardozo-Filho,

“Thermodynamic analysis of autothermal reforming of methane via entropy

maximization: Hydrogen production,” Int. J. Hydrogen Energy, vol. 39, no. 16, pp. 8257–

8270, 2014.

[21] Y. Lwin, W. R. W. Daud, A. B. Mohamad, and Z. Yaakob, “Hydrogen production from

steam-methanol reforming: Thermodynamic analysis,” Int. J. Hydrogen Energy, vol. 25,

no. 1, pp. 47–53, 2000.

[22] A. Lutz, R. Bradshaw, J. Keller, and D. Witmer, “Thermodynamic analysis of hydrogen

production by steam reforming,” Int. J. Hydrogen Energy, vol. 28, no. 2, pp. 159–167,

2003.

[23] Z.-W. Liu, K.-W. Jun, H.-S. Roh, and S.-E. Park, “Hydrogen production for fuel cells

through methane reforming at low temperatures,” J. Power Sources, vol. 111, no. 1, pp.

283–287, 2002.

104

[24] P. Ribeirinha, M. Boaventura, J. C. B. Lopes, J. M. Sousa, and A. Mendes, “Study of

different designs of methanol steam reformers: Experiment and modeling,” Int. J.

Hydrogen Energy, vol. 39, no. 35, pp. 19970–19981, 2014.

[25] A. Kumar, M. Baldea, and T. F. Edgar, “On Optimal Sensing and Actuation Design for an

Industrial Scale Methane Reformer Furnace,” AlChE J., vol. 62, no. 9, pp. 3225–3237,

2016.

[26] P. J. De Wild and M. J. F. M. Verhaak, “Catalytic production of hydrogen from

methanol,” Catal. Today, vol. 60, no. 1, pp. 3–10, 2000.

[27] P. Reuse, A. Renken, K. Haas-Santo, O. Görke, and K. Schubert, “Hydrogen production

for fuel cell application in an autothermal micro-channel reactor,” Chem. Eng. J., vol. 101,

no. 1–3, pp. 133–141, 2004.

[28] R. Y. Chein, L. C. Chen, Y. C. Chen, and J. N. Chung, “Heat transfer effects on the

methanol-steam reforming with partially filled catalyst layers,” Int. J. Hydrogen Energy,

vol. 34, no. 13, pp. 5398–5408, 2009.

[29] C. Lee, C. Shen, Y. Cheng, and Y. Chang, “Flexible Pressure , Temperature , and Flow

Microsensors for Integration in Methanol Microreformer,” Sensors Mater., vol. 25, no. 6,

pp. 411–416, 2013.

[30] S. Nagano, H. Miyagawa, O. Azegami, and K. Ohsawa, “Heat transfer enhancement in

methanol steam reforming for a fuel cell,” Energy Convers. Manag., vol. 42, no. 15–17,

pp. 1817–1829, 2001.

[31] T. Matsui, C. Saburi, S. Okuda, H. Muroyama, and K. Eguchi, “Real-time sensing of

105

methane steam reforming by YSZ oxygen sensor,” Int. J. Hydrogen Energy, vol. 36, no. 4,

pp. 2945–2949, 2011.

[32] C. Y. Lee, S. J. Lee, C. C. Shen, C. T. Yeh, C. C. Chang, and Y. M. Lo, “Application of

flexible micro temperature sensor in oxidative steam reforming by a methanol micro

reformer,” Sensors, vol. 11, no. 2, pp. 2246–2256, 2011.

[33] Omega, “RTD Resistance Temperature Detectors.” [Online]. Available:

http://www.omega.com/techref/RTD_Intro.html. [Accessed: 06-Jul-2017].

[34] Omega, “Fixed Mount Thermal Imager.” [Online]. Available:

http://www.omega.ca/pptst_eng/OSXL-101.html. [Accessed: 20-Jul-2017].

[35] M. Buric, P. Ohodnicki, A. Yan, S. Huang, and K. P. Chen, “Distributed fiber-optic

sensing in a high-temperature solid-oxide fuel cell,” Proc. SPIE, vol. 9977, p. 997708,

2016.

[36] Y.-J. Rao, “In-fibre Bragg grating sensors,” Meas. Sci. Technol., vol. 8, no. 4, pp. 355–

375, 1997.

[37] H. J. Patrick, G. M. Williams, A. D. Kersey, J. R. Pedrazzani, and A. M. Vengsarkar,

“Hybrid fiber Bragg grating/long period fiber grating sensor for strain/temperature

discrimination,” IEEE Photonics Technol. Lett., vol. 8, no. 9, pp. 1223–1225, 1996.

[38] S. Mirza and J. Shtargot, “Modern thermocouples and a high-resolution delta-sigma ADC

enable high-precision temperature measurement | EE Times,” 2011.

[39] Thermocouple info, “Type K Thermocouple - Type K Thermocouples - K Type

Thermocouples - K Type Thermocouple.” [Online]. Available:

106

http://www.thermocoupleinfo.com/type-k-thermocouple.htm. [Accessed: 20-Jul-2017].

[40] Omega, “RTD Sensors & Probes - Resistance Temperature Detection | OMEGA

Engineering.” [Online]. Available: http://www.omega.com/prodinfo/rtd.html#faq.

[Accessed: 20-Jul-2017].

[41] A. Tong, “Improving the accuracy of temperature measurements,” Sens. Rev. Res. Occup.

Stress Well-being, vol. 21, no. 12, pp. 193–198, 2001.

[42] B. A. L. Gwandu, Zhang L, Chisholm K, Y. Liu, X. Shu, and I. Bennion, “Compact FBG

array structure for high spatial resolution distributed strain sensing,” Meas. Sci. Technol,

vol. 12, pp. 918–921, 2001.

[43] M. M. Werneck, R. C. S. B. Allil, B. A. Ribeiro, and F. V. B. De Nazaré, “A Guide to

Fiber Bragg Grating Sensors,” in Current Trends in Short- and Long-periods Fiber

Gratings, Christian Cuadrado-Laborde, Ed. 2013, p. 1.

[44] K. O. Hill and G. Meltz, “Fiber Bragg Grating Technology Fundamentals and Overview,”

J. Light. Technol., vol. 15, no. 8, 1997.

[45] P. J. Lemaire, R. M. Atkins, V. Mizrahi, and W. A. Reed, “High pressure H2 loading as a

technique for achieving ultrahigh UV photosensitivity and thermal sensitivity in GeO2

doped optical fibres,” Electron. Lett., vol. 29, no. 13, p. 1191, 1993.

[46] S. J. Mihailov, “Fiber bragg grating sensors for harsh environments,” Sensors, vol. 12, no.

2, pp. 1898–1918, 2012.

[47] J. Canning, M. Stevenson, S. Bandyopadhyay, and K. Cook, “Extreme silica optical fibre

gratings,” Sensors, vol. 8, no. 10, pp. 6448–6452, 2008.

107

[48] C. Chojetzki, “High-reflectivity draw-tower fiber Bragg gratings—arrays and single

gratings of type II,” Opt. Eng., vol. 44, no. 6, p. 60503, 2005.

[49] E. Lindner, C. Chojetzki, S. Brückner, M. Becker, M. Rothhardt, and H. Bartelt, “Thermal

regeneration of fiber Bragg gratings in photosensitive fibers without hydrogen loading,”

Opt. Express, vol. 17, no. 15, pp. 12523–12531, 2009.

[50] D. Grobnic, C. Hnatovsky, and S. J. Mihailov, “Low loss Type II regenerative Bragg

gratings made with ultrafast radiation,” Opt. Express, vol. 24, no. 25, pp. 28704–28712,

2016.

[51] P. J. Lemaire, “Reliability of optical fibers exposed to hydrogen: prediction of long-term

loss increases,” Opt. Eng., vol. 30, no. 6, pp. 780–789, 1991.

[52] A. E. Barnes, R. G. May, S. Gollapudi, and R. O. Claus, “Sapphire fibers: optical

attenuation and splicing techniques,” Appl. Opt., vol. 34, no. 30, p. 6855, 1995.

[53] A. D. Kersey, M. A. Davis, H. J. Patrick, M. LeBlanc, K. P. Koo, C. G. Askins, M. A.

Putnam, and E. J. Friebele, “Fiber grating sensors,” Light. Technol. J., vol. 15, no. 8, pp.

1442–1463, 1997.

[54] J. M. Lopez-Higuera, L. Rodriguez Cobo, A. Quintela Incera, and A. Cobo, “Fiber Optic

Sensors in Structural Health Monitoring,” J. Light. Technol., vol. 29, no. 4, pp. 587–608,

2011.

[55] R. M. Measures, Structural Monitoring with Fiber Optic Technology. Academic Press,

2001.

[56] P. Moyo, J. M. W. Brownjohn, R. Suresh, and S. C. Tjin, “Development of fiber Bragg

108

grating sensors for monitoring civil infrastructure,” Eng. Struct., vol. 27, pp. 1828–1834,

2005.

[57] F. M. Haran, J. K. Rew, and P. D. Foote, “A strain-isolated fibre Bragg grating sensor for

temperature compensation of fibre Bragg grating strain sensors,” Meas. Sci. Technol., vol.

9, no. 8, pp. 1163–1166, 1998.

[58] X. Qiao and M. A. Fiddy, “Distributed optical fiber Bragg grating sensor for simultaneous

measurement of pressure and temperature in the oil and gas downhole,” Proc. SPIE - Int.

Soc. Opt. Eng., vol. 4870, no. 69877025, pp. 554–558, 2002.

[59] M. Shou, G. Jianchun, G. Yanli, C. Jun, and Z. Boming, “On-line monitoring system for

downhole temperature and pressure,” Opt. Eng., vol. 53, no. 8, p. 87102, Aug. 2014.

[60] C. Bueley and P. M. Wild, “A flexible, highly sensitive catheter for high resolution

manometry based on in-fibre Bragg gratings,” Meas. Sci. Technol., vol. 24, no. 9, p.

94003, 2013.

[61] N. A. David, P. M. Wild, J. Hu, and N. Djilali, “In-fibre Bragg grating sensors for

distributed temperature measurement in a polymer electrolyte membrane fuel cell,” J.

Power Sources, vol. 192, pp. 376–380, 2009.

[62] N. A. David, P. M. Wild, J. Jensen, T. Navessin, and N. Djilali, “Simultaneous In Situ

Measurement of Temperature and Relative Humidity in a PEMFC Using Optical Fiber

Sensors,” J. Electrochem. Soc., vol. 157, no. 8, p. B1173, 2010.

[63] J. Canning, S. Bandyopadhyay, M. Stevenson, P. Biswas, J. Fenton, and M. Aslund,

“Regenerated gratings,” J. Eur. Opt. Soc., vol. 4, 2009.

109

[64] S. Bandyopadhyay, J. Canning, M. Stevenson, and K. Cook, “Ultrahigh-temperature

regenerated gratings in boron-codoped germanosilicate optical fiber using 193 nm.,” Opt.

Lett., vol. 33, no. 16, pp. 1917–1919, 2008.

[65] J. Canning, S. Bandyopadhyay, M. Stevenson, and K. Cook, “Fiber Bragg grating sensor

for high temperature application,” in OECC/ACOFT 2008 - Joint Conference of the Opto-

Electronics and Communications Conference and the Australian Conference on Optical

Fibre Technology, 2008, pp. 1–2.

[66] E. Lindner, J. Canning, C. Chojetzki, S. Brückner, M. Becker, M. Rothhardt, and H.

Bartelt, “Post-hydrogen-loaded draw tower fiber Bragg gratings and their thermal

regeneration.,” Appl. Opt., vol. 50, no. 17, pp. 2519–2522, 2011.

[67] S. Bandyopadhyay, J. Canning, P. Biswas, M. Stevenson, and K. Dasgupta, “A study of

regenerated gratings produced in germanosilicate fibers by high temperature annealing.,”

Opt. Express, vol. 19, no. 2, pp. 1198–1206, 2011.

[68] G. Laffont, R. Cotillard, and P. Ferdinand, “Multiplexed regenerated fiber Bragg gratings

for high-temperature measurement,” Meas. Sci. Technol., vol. 24, no. 9, p. 94010, 2013.

[69] G. Laffont, R. Cotillard, P. Ferdinand, J.-P. Jeannot, and G. Rodriguez, “Regenerated fiber

Bragg grating sensors for high temperature monitoring in sodium-cooled fast reactor,” Int.

Conf. Fast React. Relat. Fuel Cycles, pp. 1–10, 2013.

[70] D. Barrera, V. Finazzi, J. Villatoro, S. Sales, and V. Pruneri, “Performance of a high-

temperature sensor based on regenerated fiber Bragg gratings,” Proc. SPIE, vol. 7753,

2011.

110

[71] D. Barrera, V. Finazzi, J. Villatoro, S. Sales, and V. Pruneri, “Packaged optical sensors

based on regenerated fiber bragg gratings for high temperature applications,” IEEE Sens.

J., vol. 12, no. 1, pp. 107–112, 2012.

[72] R. B. Walker, S. Yun, H. Ding, M. Charbonneau, D. Coulas, P. Lu, S. J. Mihailov, N.

Ramachandran, C. Canada, and F. P. Group, “High resolution temperature mapping of gas

turbine combustor simulator exhaust with femtosecond laser induced fiber Bragg

gratings,” Proc. SPIE, vol. 10110, 2017.

[73] M. Willsch, T. Bosselmann, P. Flohr, R. Kull, W. Ecke, I. Latka, D. Fischer, and T. Thiel,

“Design of fiber optical high temperature sensors for gas turbine monitoring,” Proc. SPIE,

vol. 7503, 2009.

[74] R. J. Black, J. M. Costa, B. Moslehi, L. Zarnescu, D. Hackney, and K. Peters, “Fiber optic

temperature profiling for thermal protection heat shields,” Proc. SPIE, vol. 9062, 2014.

[75] A. Khosravi, “The development of a metal plate test reactor for studying reaction kinetics

on catalytically coated heat transfer components,” Queen’s University, 2012.

[76] Flexitallic, “Thermiculite 866/866 LS.”

[77] S. D. Angeli, G. Monteleone, A. Giaconia, and A. A. Lemonidou, “State-of-the-art

catalysts for CH4 steam reforming at low temperature,” Int. J. Hydrogen Energy, vol. 39,

no. 5, pp. 1979–1997, 2014.

111

Appendix A – Technical Drawings

This appendix includes the technical drawing and section view of the reformer to

understand the modifications that were made to accommodate the fiber.

Figure A.1 - Technical drawing for the modification of the reformer base.

112

Figure A.2 - Section view of the gas inlet end of the reformer base showing the modification

needed to accommodate the fiber.

113

Appendix B – Characterization values

B.1. Methanol steam reforming

Table B-1 contains the values for the five characterization runs performed on the FBG array used to monitor temperature during the

methanol steam reforming trials as well as the average calculated from these five instances.

Table B-1 - Characterization values for the five characterization runs and the average for each FBG

Run 1 Run 2 Run 3 Run 4 Run 5 Average

FBG 1 λ0 (nm) 1529.067 1529.078 1529.082 1529.081 1529.100 1529.082

KT (pm/°C) 13.9 13.9 13.8 13.8 13.7 13.8

FBG 2 λ0 (nm) 1534.038 1534.061 1534.076 1534.073 1534.095 1534.067

KT (pm/°C) 14.1 14.0 13.9 13.9 13.8 13.9

FBG 3 λ0 (nm) 1538.982 1538.997 1539.010 1539.009 1539.029 1539.005

KT (pm/°C) 14.0 14.0 13.9 13.9 13.8 13.9

FBG 4 λ0 (nm) 1543.996 1544.010 1544.030 1544.028 1544.052 1544.023

KT (pm/°C) 14.2 14.1 14.0 14.0 13.9 14.0

114

FBG 5 λ0 (nm) 1548.937 1548.956 1548.977 1548.977 1548.997 1548.964

KT (pm/°C) 14.2 14.1 14.0 14.0 13.9 14.0

FBG 6 λ0 (nm) 1553.968 1553.987 1554.010 1554.009 1554.029 1554.001

KT (pm/°C) 14.3 14.2 14.2 14.1 14.1 14.2

FBG 7 λ0 (nm) 1558.859 1558.877 1558.901 1558.903 1558.921 1558.892

KT (pm/°C) 14.3 14.2 14.1 14.1 14.0 14.1

Figure B.1 illustrates the relation between wavelength and temperature for all seven FBGs during characterization Run 1.

115

Figure B.1 – Wavelength as a function of temperature for FBG 1-7 during characterization Run

1.

116

B.2. Methane steam reforming

Table B-2 - Characterization values for the five pre-regeneration characterization runs and the average for each FBG on the FBG array


FBG 1 λ0 (nm) 1529.005 1529.053 1529.042 1529.041 1529.028 1529.034

KT (pm/°C) 14.1 13.9 13.9 13.9 13.9 13.9

FBG 2 λ0 (nm) 1533.994 1534.032 1534.019 1534.019 1534.006 1534.014

KT (pm/°C) 14.0 13.8 13.9 13.9 13.9 13.9

FBG 3 λ0 (nm) 1538.937 1538.976 1538.967 1538.959 1538.949 1538.958

KT (pm/°C) 14.1 13.9 13.9 14.0 14.0 14.0

FBG 4 λ0 (nm) 1543.993 1544.039 1544.024 1544.012 1544.003 1544.014

KT (pm/°C) 14.2 14.0 14.0 14.1 14.1 14.1

FBG 5 λ0 (nm) 1548.960 1548.993 1548.989 1548.972 1548.963 1548.975

KT (pm/°C) 14.2 14.1 14.1 14.1 14.2 14.1

FBG 6 λ0 (nm) 1553.953 1553.995 1553.994 1553.975 1553.964 1553.976

KT (pm/°C) 14.3 14.1 14.1 14.2 14.2 14.2

117

FBG 7 λ0 (nm) 1558.907 1558.958 1558.951 1558.938 1558.925 1558.936

KT (pm/°C) 14.3 14.2 14.2 14.2 14.3 14.2

Table B-3 - Pre-regeneration characterization values obtained from 230 to 290 °C for the single FBG


λ0 (nm) 1534.405 1534.421 1534.402 1534.413 1534.412 1534.411

KT (pm/°C) 13.5 13.5 13.6 13.5 13.5 13.5

Table B-4 - Post-regeneration characterization values for the single FBG

Characterization Run 1 Run 2 Run 3 Run 4 Average

300 °C to 900 °C λ0 (nm) 1532.820 1532.600 1532.675 1532.710 1532.701

KT (pm/°C) 15.0 15.3 15.2 15.2 15.2

500 °C to 900 °C λ0 (nm) 1532.402 1532.417 1532.408 1532.413 1532.410

KT (pm/°C) 15.6 15.6 15.6 15.6 15.6

700 °C to 900 °C λ0 (nm) 1531.991 1531.996 1532.034 1532.116 1532.027

KT (pm/°C) 16.1 16.1 16.0 15.9 16.0

118

Appendix C – Data analysis

C.1. Stabilisation period and baseline

The first step to analyse the data for both the methanol and methane conversion percentage is

to establish the baseline from which further changes in temperature will be compared to plot the

graphs of temperature change as a function of position in the reformer. While the following

procedure is repeated for each furnace temperature and each FBG, as evident from Table C-1 and

Table C-2, the following data analysis example is presented for FBG 1 at a furnace temperature of

250 °C. Figure C.1 illustrates the wavelength of FBG 1 during the stabilisation period. This

stabilisation period lasts ~15 hours and is used to ensure the temperature of the reformer is at

steady state.

By looking at Figure C.1, it is possible to see that there is variation in the wavelength of FBG

1. These are caused by variation in room temperature during the stabilisation period which is done

overnight. Averaging over this entire period would not give us an accurate baseline. Therefore, the

average wavelength over the last 5 minutes of the stabilisation period is calculated and used as

baseline for this FBG at this furnace temperature. Immediately after the end of the stabilisation

period, the flow rate is increased from 0.010 ml/min to the desired test flow rates. Table C-1

contains the baseline value for each FBG at each of the three furnace temperatures for the methanol

steam reforming experiments while Table C-2 contains these values for the methane steam

reforming experiments.

119

Figure C.1 - Wavelength over time for FBG 1 during the stabilisation period for the furnace


Table C-1 - FBG baseline values for the methanol steam reforming experiments

Furnace

temperature

FBG 1

(nm)

FBG 2

(nm)

FBG 3

(nm)

FBG 4

(nm)

FBG 5

(nm)

FBG 6

(nm)

FBG 7

(nm)

250 °C 1532.353 1537.361 1542.300 1547.346 1552.296 1557.351 1536.213

265 °C 1532.554 1537.561 1542.502 1547.548 1552.496 1557.553 1562.414

280 °C 1532.746 1537.753 1542.693 1547.740 1552.695 1557.750 1562.614

Table C-2 - FBG baseline values for the methane steam reforming experiments

Furnace

temperature

FBG 1

(nm)

FBG 2

(nm)

FBG 3

(nm)

FBG 4

(nm)

FBG 5

(nm)

FBG 6

(nm)

FBG 7

(nm)

650 °C 1537.493 1542.406 1547.396 1552.606 1557.680 1562.651 1567.497

750 °C 1539.513 1544.367 1549.358 1554.587 1559.723 1564.647 1569.472

775 °C 1540.163 1544.854 1549.854 1555.028 1560.150 1565.039 1569.906

120

The same procedure detailed here for the FBG is used to obtain the baseline for the

thermocouples. Table C-3 and Table C-4 present the baseline values for the thermocouples for the

methanol and methane steam reforming experiments respectively.

Table C-3 - Thermocouple baseline values for the methanol steam reforming experiments

Furnace

temperature

Thermo. 1

(°C)

Thermo. 2

(°C)

Thermo. 3

(°C)

Thermo. 4

(°C)

Thermo. 5

(°C)

Thermo. 6

(°C)

250 °C 244.1 244.3 246.6 244.9 245.6 241.7

265 °C 258.2 258.5 260.9 259.2 268.3 256.0

280 °C 272.2 272.6 274.9 273.3 282.6 269.9

Table C-4 - Thermocouple baseline values for the methane steam reforming experiments

Furnace

temperature

Thermo. 1

(°C)

Thermo. 2

(°C)

Thermo. 3

(°C)

Thermo. 4

(°C)

Thermo. 5

(°C)

Thermo. 6

(°C)

650 °C 625.5 637.1 632.4 635.3 630.3 626.1

750 °C 726.0 739.4 734.7 738.0 733.9 727.8

775 °C 751.5 765.1 760.3 763.8 759.7 753.1

C.2. Temperature change measurement

This section of Appendix will explain how the data is processed to obtain the temperature change

measurements presented in the graphs of Chapter 4 and 5. The example given here uses the FBG

1 at a furnace temperature of 250 °C and at a flow rate of 0.070 ml/min. It is important to note that

121

this data analysis procedure was used for both the methanol and methane steam reforming

experiments.

During the steam reforming experiments, each flow rate is maintained for approximately

one and a half hour. This ensure that there is enough time for the conditions in the reformer to

reach steady state and for the gas chromatograph to analyze two different samples. The gas

chromatograph samples the gas exiting the reformer at 30 minutes interval. Figure C.2 illustrates

the wavelength of FBG 1 during the experiments at a furnace temperature of 250 °C and a flow

rate of 0.070 ml/min.

122

Figure C.2 - Wavelength of FBG 1 over time during the methanol steam reforming experiments

at a furnace temperature of 250 °C and a flow rate of 0.070 ml/min.

As can be seen in Figure C.2, once the flow rate is increased (time=0 hr), it takes some time

for the temperature in the reformer to adjust. Taking the average wavelength over the whole period

at which the flow rate is 0.070 ml/min would not accurately represent the temperature change in

the reformer. For this reason, the temperature change in the reformer as seen in Chapter 4 and 5 is

obtained with the average wavelength over a one minute period during the experiments at a given

flow rate, 0.070 ml/min for the example given here. The following procedure was followed at

every flow rate. This procedure was established to ensure a systematic way to obtain the

temperature change measurement for each flow rate. It also ensured that the data selected for the

123

temperature change graph was collected during a phase of steady state regime in the reformer. The

data analysis procedure is as follows;

1. For every data point recorded, the average for the previous 10 minutes is calculated. This

will now be referred to as the 10 minutes average.

2. Find the maximum and minimum 10 minutes average and calculate the difference

between the maximum and the minimum.

3. Calculate the percentage of relative difference between the 10 minutes average of a data

point at time t and the 10 minute average of the data point at time t-1.

% 𝑟𝑒𝑙𝑎𝑡𝑖𝑣𝑒 𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒 = |10 𝑚𝑖𝑛𝑢𝑡𝑒𝑠 𝑎𝑣𝑒𝑟𝑎𝑔𝑒𝑡𝑖𝑚𝑒 𝑡 − 10 𝑚𝑖𝑛𝑢𝑡𝑒𝑠 𝑎𝑣𝑒𝑟𝑎𝑔𝑒𝑡𝑖𝑚𝑒 𝑡−1|

𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑏𝑒𝑡𝑤𝑒𝑒𝑛 𝑚𝑎𝑥𝑖𝑚𝑢𝑚 𝑎𝑛𝑑 𝑚𝑖𝑛𝑖𝑚𝑢𝑚 𝑐𝑎𝑙𝑐𝑢𝑙𝑎𝑡𝑒𝑑 𝑖𝑛 2∗ 100

4. Sum the percentage relative difference calculated for all 7 FBGs.

5. Find the minimum value from the sums calculated in 4.

6. The time at which the minimum value found in step 5 happens will be the starting point

for the one minute period of data which is used for the flow rate of 0.070 ml/min.

7. The average wavelength over this one minute period is calculated.

8. The average wavelength from step 7 is subtracted from to the baseline wavelength for

FBG 1 found in Appendix C.1.

9. Using the temperature sensitivity of FBG 1, the wavelength change calculated in step 8 is

converted to a temperature change.

The temperature change found in step 9 is the temperature change for FBG 1 at a furnace

temperature of 250 °C and a flow rate of 0.070 ml/min found in Figure 13.

124

The time at which the minimum value is found in step 5 is also used as the time for which the

temperature change measured by the thermocouple is taken. This ensure that in the graphs of

Chapter 4 and Chapter 5, the thermocouple and FBG measurements are taken simultaneously. Step

7, 8 and 9 are repeated with the thermocouple data and the temperature change found in step 9 is

the temperature change for Thermocouple 1 at a furnace temperature of 250 °C and a flow rate of

0.070 ml/min found in Figure 16.

125

Appendix D – Uncertainty

This appendix presents the uncertainty values for the methanol and methane steam reforming trials.

These values are obtained by calculation the standard deviation of the date during the on minute

period of data collection, as described in Appendix C. All uncertainty values are in °C.

Table D-1 - Uncertainty values for the FBGs, methanol steam reforming at a furnace temperature

of 250 °C

Position/Flow

rate


0.07 ml/min 0.1 0.1 0.2 0.2 0.2 0.2 0.3

0.09 ml/min 0.1 0.2 0.1 0.1 0.1 0.2 0.2

0.11 ml/min 0.1 0.2 0.2 0.2 0.2 0.2 0.2

0.13 ml/min 0.1 0.2 0.2 0.2 0.2 0.2 0.2

Table D-2 - Uncertainty values for the thermocouples, methanol steam reforming at a furnace

temperature of 250 °C

Position/Flow

rate

Thermo.

1

Thermo.

2

Thermo.

3

Thermo.

4

Thermo.

5

Thermo.

6

0.07 ml/min 0.1 0.1 > 0.1 > 0.1 > 0.1 0.1

0.09 ml/min 0.1 > 0.1 0.1 > 0.1 0.1 0.1

0.11 ml/min 0.1 > 0.1 0.0 > 0.1 0.1 > 0.1

0.13 ml/min 0.1 0.1 0.0 > 0.1 > 0.1 0.1


of 265 °C

Position/Flow

rate


0.07 ml/min 0.1 0.2 0.2 0.1 0.2 0.2 0.3

0.09 ml/min 0.1 0.3 0.2 0.1 0.2 0.1 0.4

0.11 ml/min 0.1 0.3 0.2 0.2 0.2 0.2 0.4

126



Position/Flow

rate

Thermo.

1

Thermo.

2

Thermo.

3

Thermo.

4

Thermo.

5

Thermo.

6

0.07 ml/min > 0.1 > 0.1 > 0.1 > 0.1 > 0.1 > 0.1

0.09 ml/min 0.1 0.1 0.1 > 0.1 0.1 > 0.1

0.11 ml/min 0.1 0.1 > 0.1 > 0.1 0.1 0.1


of 280 °C

Position/Flow

rate


0.13 ml/min 0.2 0.2 0.2 0.3 0.2 0.2 0.2

0.26 ml/min 0.2 0.2 0.2 0.3 0.2 0.2 0.2

0.32 ml/min 0.2 0.3 0.2 0.2 0.2 0.2 0.3



Position/Flow

rate

Thermo.

1

Thermo.

2

Thermo.

3

Thermo.

4

Thermo.

5

Thermo.

6

0.13 ml/min 0.1 0.1 0.1 > 0.1 > 0.1 0.1

0.26 ml/min > 0.1 0.1 > 0.1 > 0.1 0.1 0.1

0.32 ml/min 0.1 0.1 0.1 0.1 0.1 0.1

Table D-7 - Uncertainty values for the FBGs, methane steam reforming at a furnace temperature

of 650 °C

Position/Flo

w rate


25 ml/min 0.1 0.1 0.1 0.1 0.1 0.1 0.1

40 ml/min 0.1 0.1 0.1 0.1 0.1 0.1 0.2

60 ml/min 0.1 0.1 0.1 0.1 0.1 0.1 0.2

90 ml/min 0.1 0.1 0.1 0.1 0.1 0.1 0.1

127

Table D-8 - Uncertainty values for the thermocouples, methane steam reforming at a furnace


Position/Flow

rate

Thermo.

1

Thermo.

2

Thermo.

3

Thermo.

4

Thermo.

5

Thermo.

6

25 ml/min > 0.1 > 0.1 > 0.1 > 0.1 > 0.1 0.1

40 ml/min > 0.1 > 0.1 > 0.1 0.1 > 0.1 0.1

60 ml/min > 0.1 > 0.1 > 0.1 > 0.1 > 0.1 > 0.1

90 ml/min > 0.1 > 0.1 > 0.1 0.1 > 0.1 0.1

Table D-9 - Uncertainty values for the FBGs, methane steam reforming at a furnace temperature

of 750 °C

Position/Flo

w rate


25 ml/min 0.1 0.1 0.1 0.1 0.2 0.2 0.5

40 ml/min 0.1 0.2 0.1 0.2 0.1 0.2 0.1

60 ml/min 0.1 0.1 0.1 0.1 0.2 0.1 0.1

90 ml/min 0.1 0.1 0.1 0.1 0.2 0.2 0.1

Table D-10 - Uncertainty values for the thermocouples, methane steam reforming at a furnace


Position/Flow

rate

Thermo.

1

Thermo.

2

Thermo.

3

Thermo.

4

Thermo.

5

Thermo.

6

25 ml/min > 0.1 > 0.1 > 0.1 > 0.1 > 0.1 0.1

40 ml/min > 0.1 > 0.1 > 0.1 0.1 > 0.1 0.1

60 ml/min > 0.1 > 0.1 > 0.1 > 0.1 > 0.1 > 0.1

90 ml/min > 0.1 > 0.1 > 0.1 > 0.1 > 0.1 0.1

Table D-11 - Uncertainty values for the FBGs, methane steam reforming at a furnace


Position/Flo

w rate


20 ml/min 0.1 0.1 0.1 0.2 0.1 0.3 0.9

25 ml/min 0.1 0.1 0.1 0.2 0.1 0.2 0.7

30 ml/min 0.1 0.1 0.1 0.2 0.1 0.3 0.7

40 ml/min 0.1 0.1 0.1 0.1 0.1 0.2 0.1

60 ml/min 0.1 0.1 0.1 0.1 0.1 0.2 0.1

128

Table D-12 - Uncertainty values for the FBGs, methane steam reforming at a furnace


Position/Flow

rate

Thermo.

1

Thermo.

2

Thermo.

3

Thermo.

4

Thermo.

5

Thermo.

6

20 ml/min > 0.1 > 0.1 0.1 0.1 > 0.1 0.1

25 ml/min > 0.1 > 0.1 > 0.1 > 0.1 > 0.1 0.1

30 ml/min > 0.1 > 0.1 > 0.1 > 0.1 > 0.1 0.1

40 ml/min > 0.1 > 0.1 > 0.1 0.1 > 0.1 0.1

60 ml/min > 0.1 > 0.1 > 0.1 0.1 > 0.1 0.1

129

Appendix E – Methanol Conversion Percentage

This appendix provides a sample calculation for the methanol conversion percentage, the

numbers used are for the experiments done at a furnace temperature of 250 °C and a flow rate of

0.070 ml/min.

𝑋𝐶𝐻3𝑂𝐻 =𝐹𝐻2

+ 𝐹𝐶𝑂

3𝐹𝐶𝐻3𝑂𝐻

XCH3OH: methanol conversion percentage

Fi: molar flow rate of component i (mol/s)

yi: normalized fraction of component i

VF: flow rate of dry product gas (ml/s)

𝐹𝑖 = 𝑦𝑖 ∗ 𝑉𝐹 ∗𝑃𝑜

𝑅𝑇𝑜= 𝑦𝑖 ∗ 𝑉𝐹 ∗ 4.4615 ∗ 10−5 [𝑚𝑜𝑙/𝑠]

The volumetric flow rate of the dry product gas is calculated by averaging the

measurements taken with the soap film meter for each reactant flow rates. Five measurements were

taken at each flow rates tested. For a (reactant) flow rate of 0.07 ml/min, the dry product gas flow

rate is:

𝑉𝐹 = 1.839 𝑚𝑙/𝑠

The next step is to calculate percentage of each component in the sample measured by the gas

chromatograph. The calibration equations needed for this are taken from Aida Khosravi’s thesis

[75]. Using hydrogen as an example, the conversion from the gas chromatograph reading to the

130

percentage of the component is 1/7.2192. The gas chromatograph integrated peak area for

hydrogen is 484.547 mV.s. The hydrogen percentage is:

𝐻2% = 484.547

7.2192= 67.1%

The same is done for all other components, carbon monoxide, carbon dioxide and nitrogen.

For the experiments at a furnace temperature of 250 °C and a flow rate of 0.07 ml/min, these are

1.1%, 26.1% and 0.8 % respectively. Adding the various fraction of each component as measured

by the gas chromatograph does not equal 100 % due to the calibration. Therefore, the next step is

to calculate the normalized fraction of component i in the dry product.

𝑡𝑜𝑡𝑎𝑙 % = 67.1 + 1.1 + 26.1 + 0.8 = 95.1 %

𝑦𝐻2=

𝐻2%

𝑡𝑜𝑡𝑎𝑙 %=

67.1

95.1= 0.7

It is now possible to calculate the molar flow rate of hydrogen:

𝐹𝐻2= 𝑦𝐻2

∗ 𝑉𝐹 ∗𝑃𝑜

𝑅𝑇𝑜= 𝑦𝐻2

∗ 𝑉𝐹 ∗ 4.4615 ∗ 10−5 = 0.7 ∗ 1.839 ∗ 4.4615 ∗ 10−5

= 5.74 ∗ 10−5 𝑚𝑜𝑙/𝑠

The same calculation for carbon monoxide gives a result of 9.62*10-7 mol/s. Now that the

molar flow rate of hydrogen and carbon monoxide are known, we must calculate the methanol

molar flow rate. This is done with the following equation.

131

𝐹𝐶𝐻3𝑂𝐻 = 𝜌𝐹𝑄𝐹𝑤𝐶𝐻3𝑂𝐻

𝑀𝐶𝐻3𝑂𝐻

ρF: density of the feed solution (g/cm3)

QF: volumetric flow rate of the feed solution (cm3/min)

wCH3OH: weight fraction of the methanol in the feed solution based on the density of

the feed solution

MCH3OH: molecular weight of methanol (32.02 g/mol)

ρH2O: density of water (0.9982 g/cm3)

𝐷2020: specific gravity of the feed solution

The density of the feed mixture is calculated based on the density of water and the specific

gravity of the feed solution. The specific gravity of the feed solution is calculated based on the

weight of the feed solution divided by the weight of the same volume of deionized water. This

value is 0.89 [75].

𝜌𝐹 = 𝜌𝐻2𝑂 ∗ 𝐷2020 = 0.9982 ∗ 0.89 = 0.89 𝑔/𝑐𝑚3

The flow rate of the feed pump in this example is 0.070 ml/min or 0.0012 ml/s. The weight

fraction of the methanol in the feed solution based on the density of the feed solution is 0.6290

[75].

𝐹𝐶𝐻3𝑂𝐻 = 𝜌𝐹𝑄𝐹𝑤𝐶𝐻3𝑂𝐻

𝑀𝐶𝐻3𝑂𝐻=

0.89 ∗ 0.0012 ∗ 0.6290

32.02= 2.04 ∗ 10−5 𝑚𝑜𝑙/𝑠

𝑋𝐶𝐻3𝑂𝐻 =𝐹𝐻2

+ 𝐹𝐶𝑂

3𝐹𝐶𝐻3𝑂𝐻=

5.74 ∗ 10−5 + 9.62 ∗ 10−7

3 ∗ 2.04 ∗ 10−5= 0.94.1 = 94.1 %

132

Appendix F – Regeneration

As mentioned briefly in Chapter 5, the regeneration of another array was done in-situ but failed

before the sensors could be used to monitor the temperature in the methane steam reforming. The

present appendix presents the results from this regeneration. The regeneration procedure used for

the methane steam reforming experiments was described in Chapter 3. The graph in Figure F.1

illustrates the progress of the regeneration process through the power of the FBG and the

temperature over time. In this graph, the time zero represents the time at which the regeneration

process was begun, when the temperature was set to increase to the first intermediate step as

described in Chapter 3.

Figure F.1- Power and temperature over time during the regeneration of the FBG array

133

In Figure F.2, it is possible to observe the regeneration process over time by seeing the power

of each FBG decrease until it reaches the noise floor, 0 dB, and then regenerate. Shortly after the

regeneration of FBG 5, 6 and 7 was complete, these regenerated FBGs showed signs of instability.

In Figure F.2, this is observable for FBG 5. As seen in this graph, the signal of FBG 5 started

erasing again after regeneration, at approximately five hours. In this thesis, this is referred to as

second erasure. The first FBG to show instability was FBG 7 shortly followed by FBG 6. There

was a complete erasure of the grating and no subsequent regeneration was observed.

In an attempt to stabilize the array and prevent the loss of more FBGs, the temperature was

then decreased from 910 °C to 700 °C which is the last temperature step that can be seen in Figure

F.1.

The graphs in Figure F.2 through Figure F.4 illustrate the secondary erasure with the help of

spectrum of the FBG array at different points in time. A time of 0:00 corresponds to the beginning

of the regeneration process as defined for Figure F.1. To facilitate data analysis, the FBGs are

located on the fiber, and named, in order of increasing Bragg wavelength. Therefore, FBG 1 is

located on the left of the graph and has an initial Bragg wavelength of ~1530 nm and FBG 7, ~1560

nm, is the last grating on the right side of the graph.

134

Figure F.2 - Spectrum of the FBG array showing the secondary erasure of FBG 6 and 7.

In Figure F.2, secondary erasure can be observed for the first time. FBGs 6 and 7 are already

regenerated. As the temperature is kept constant to allow for complete regeneration of FBG 1, 2,

3 and 4, the signal of FBG 6 and 7 is becomes weaker between 4:45 and 4:50. This leads to

secondary erasure of these two FBGs. In Figure F.3, another example of secondary erasure can be

observed, this time for FBG 5.

135

Figure F.3 - Spectrum of the FBG array showing the secondary erasure of FBG 5.

The graph in Figure F.3 illustrates the secondary erasure of FBG 5 which occurs between

5:12 and 5:18. This figure also shows that FBG 1 through 4 are fully regenerated and stable.

Following this, the temperature was dropped to 700 °C in an attempt to stabilize and retain FBGs

1 through 4.

Unfortunately, after spending 24 hours between 650 °C and 750 °C and a subsequent 48 hours

at 350 °C, FBG 4 also went through secondary erasure. The data for the secondary erasure of FBG

4 is not available. At this time, the furnace was shut down and the reformer allowed to cool from

136

350 °C to room temperature over a period of 15 hours. During this time, FBGs 1, 2 and 3 all went

through simultaneous secondary erasure. This is presented below in Figure F.4.

Figure F.4 - Spectrum of the FBG array showing the secondary erasure of FBG 1, 2 and 3.

Since all FBGs underwent secondary erasure, the fiber was removed from the reformer and

replaced by a new identical array. This new array was regenerated to prepare the sensors for the

methane steam reforming environment. This is described in Chapter 5.

137

Appendix G – Drift Correction Factor

In Chapter 4, it was mentioned that despite the stabilisation period, the temperature in the

reformer had not reach steady state when the methane steam reforming experiments were begun.

In the graph of Figure 24, an example of the temperature drift is presented with the wavelength

drift during the experiment at a furnace temperature of 650 °C. Since there is a temperature drift

in the reformer after the baseline wavelength is measured, see Appendix C.1, the temperature

change in the reformer during the methane steam reforming experiments would not accurately

represent the effect of the endothermic chemical reaction taking place. To solve this, a drift

correction factor was applied to the data in Chapter 5. The principal assumption made in the

calculation of the drift correction factor is that the reformer uniformly drifts to higher temperatures

during the experiments. Since the temperature in the reformer does not stabilise throughout the

day, a new drift correction factor must be calculated for each methane flow rate tested. A sample

calculation is presented here for a furnace temperature of 775 °C and a methane flow rate of 20

ml/min. First the calculation of the drift correction factor will be explained in steps. It will then be

followed by a table containing the values required to calculate the drift correction factor for the

furnace temperature of 775 °C and the methane flow rate of 20 ml/min.

1. The average wavelength over the one minute of data used to calculate the temperature

change in the reformer, refer to Appendix C, is calculated for each FBG.

2. For each FBG, the baseline wavelength of this FBG for a furnace temperature of 775 °C is

subtracted from the average wavelength calculated in Step 1. This gives us the wavelength

drift in the reformer measured by each FBG for this flow rate.

138

3. The wavelength drift of each FBG calculated in step 2 is converted to a temperature drift

using the temperature sensitivity value of 15 pm/°C.

4. The average temperature drift in the reformer is calculated by averaging the seven values

obtained in step 3. This is the drift correction factor for the flow rate of 20 ml/min, furnace

temperature=775 °C.

5. To apply the drift correction factor, the temperature change measured by each FBG is

calculated according to the procedure detail in Appendix C. The drift correction factor is

then subtracted from the temperature change measured according to Appendix C.

The following table provides the sample calculation for the drift correction factor for the data

collected at a furnace temperature of 775 °C and a methane flow rate of 20 ml/min. Note that the

sample calculation and procedure detailed in this Appendix refer to the thermocouple but the

procedure followed to calculate the drift correction factor for the thermocouples is the same.

139

Table G-1 - Sample calculation for the drift correction factor for a furnace temperature of 775 °C and a methane flow rate of 20ml/mi


1 minute

average

wavelength

(Step 1) (nm)

1540.216 1544.863 1549.868 1555.055 1560.177 1565.086 1569.952

Baseline

Wavelength

(nm)

1540.163 1544.854 1549.854 1555.028 1560.150 1565.039 1569.906

Wavelength

drift (Step 2)

(nm)

0.053 0.009 0.014 0.027 0.027 0.047 0.046

Temperature

drift (Step 3)

(°C)

3.53 0.60 0.93 1.80 1.80 3.13 3.07

Drift

correction

factor (Step 4)

(°C)

2.12

Temperature

change

(Appendix C)

(°C)

3.53 0.60 0.93 1.80 1.80 3.13 3.07

Temperature

change (Step

5) (°C)

1.41 -1.52 -1.19 -0.32 -0.32 1.01 0.95

Fiber Bragg Gratings for Temperature Monitoring in ...

Documents

Transcript of Fiber Bragg Gratings for Temperature Monitoring in ...